{"api":{"version":"1.2.0","contract_url":"https://github.com/DataViking-Tech/SynthBench/blob/main/docs/leaderboard-json-api.md"},"generated_at":"2026-07-14T23:20:52.878220+00:00","synthbench_version":"0.1.0","datasets":["globalopinionqa","opinionsqa","subpop"],"entries":[{"rank":1,"config_id":"ensemble--3-model-blend--tdefault--tplcurrent--3ba7cfb3","provider":"SynthPanel Ensemble (3-model)","model":"SynthPanel Ensemble (3-model)","provider_id":"ensemble","model_id":"3-model-blend","dataset":"globalopinionqa","framework":"product","sps":0.81346,"p_dist":0.807219,"p_rank":0.68741,"p_refuse":0.945752,"jsd":0.192781,"tau":0.374819,"n":100,"ci_lower":0.785316,"ci_upper":0.842114,"is_baseline":false,"is_ensemble":true,"topic_scores":{"Economy & Work":0.689455,"General Attitudes":0.795661,"Health & Science":0.908908,"International Relations & Security":0.742844,"Politics & Governance":0.804461,"Technology & Digital Life":0.743427,"Trust & Wellbeing":0.52049},"topic_metrics":{"Economy & Work":{"sps":0.689455,"n":5,"p_dist":0.785576,"p_rank":0.593333,"p_refuse":0.90915},"General Attitudes":{"sps":0.795661,"n":11,"p_dist":0.842045,"p_rank":0.749277,"p_refuse":0.963636},"Health & Science":{"sps":0.908908,"n":2,"p_dist":0.891039,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.742844,"n":50,"p_dist":0.800787,"p_rank":0.684901,"p_refuse":0.955417},"Politics & Governance":{"sps":0.804461,"n":22,"p_dist":0.868445,"p_rank":0.740478,"p_refuse":0.903854},"Technology & Digital Life":{"sps":0.743427,"n":3,"p_dist":0.893212,"p_rank":0.593642,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.52049,"n":7,"p_dist":0.560667,"p_rank":0.480313,"p_refuse":0.981501}},"demographic_scorecard":null,"cost_usd":0.537,"cost_per_100q":0.537,"cost_per_sps_point":0.660143,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.761016,"sps_private":0.698738,"sps_public_private_delta":0.062278,"verification_badge":"flagged","run_count":2,"dataset_coverage_count":3},{"rank":2,"config_id":"synthpanel--gpt-4o-mini--tdefault--tplcurrent--ba37570c","provider":"SynthPanel (GPT-4o-mini)","model":"SynthPanel (GPT-4o-mini)","provider_id":"synthpanel","model_id":"openai/gpt-4o-mini","dataset":"globalopinionqa","framework":"product","sps":0.786229,"p_dist":0.689363,"p_rank":0.693591,"p_refuse":0.975733,"jsd":0.310637,"tau":0.387182,"n":100,"ci_lower":0.757181,"ci_upper":0.813287,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.620677,"General Attitudes":0.73761,"Health & Science":0.859111,"International Relations & Security":0.696611,"Politics & Governance":0.698229,"Technology & Digital Life":0.664402,"Trust & Wellbeing":0.57537},"topic_metrics":{"Economy & Work":{"sps":0.620677,"n":5,"p_dist":0.681686,"p_rank":0.559669,"p_refuse":0.966928},"General Attitudes":{"sps":0.73761,"n":11,"p_dist":0.732894,"p_rank":0.742325,"p_refuse":1},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.696611,"n":50,"p_dist":0.692814,"p_rank":0.700407,"p_refuse":0.973791},"Politics & Governance":{"sps":0.698229,"n":22,"p_dist":0.691145,"p_rank":0.705313,"p_refuse":0.967561},"Technology & Digital Life":{"sps":0.664402,"n":3,"p_dist":0.758977,"p_rank":0.569827,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.57537,"n":7,"p_dist":0.537183,"p_rank":0.613558,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0.063,"cost_per_100q":0.063,"cost_per_sps_point":0.080129,"cost_per_response":0.000021,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.696961,"sps_private":0.672033,"sps_public_private_delta":0.024928,"verification_badge":"verified","run_count":3,"dataset_coverage_count":3},{"rank":3,"config_id":"raw--gemini-2.5-flash--tdefault--tplcurrent--d5695573","provider":"Gemini 2.5 Flash","model":"Gemini 2.5 Flash","provider_id":"raw","model_id":"google/gemini-2.5-flash","dataset":"globalopinionqa","framework":"raw","sps":0.770427,"p_dist":0.687026,"p_rank":0.644521,"p_refuse":0.979733,"jsd":0.312974,"tau":0.289042,"n":100,"ci_lower":0.740462,"ci_upper":0.798893,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.826171,"General Attitudes":0.750929,"Health & Science":0.859111,"International Relations & Security":0.668434,"Politics & Governance":0.64137,"Technology & Digital Life":0.866699,"Trust & Wellbeing":0.333731},"topic_metrics":{"Economy & Work":{"sps":0.826171,"n":5,"p_dist":0.813397,"p_rank":0.838946,"p_refuse":0.966928},"General Attitudes":{"sps":0.750929,"n":11,"p_dist":0.792729,"p_rank":0.70913,"p_refuse":1},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.668434,"n":50,"p_dist":0.697365,"p_rank":0.639503,"p_refuse":0.979791},"Politics & Governance":{"sps":0.64137,"n":22,"p_dist":0.66963,"p_rank":0.61311,"p_refuse":0.972106},"Technology & Digital Life":{"sps":0.866699,"n":3,"p_dist":0.830239,"p_rank":0.90316,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.333731,"n":7,"p_dist":0.320273,"p_rank":0.34719,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0.0708,"cost_per_100q":0.0708,"cost_per_sps_point":0.091897,"cost_per_response":0.000047,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.679705,"sps_private":0.616382,"sps_public_private_delta":0.063323,"verification_badge":"flagged","run_count":1,"dataset_coverage_count":3},{"rank":4,"config_id":"raw--llama-3.3-70b-instruct--tdefault--tplcurrent--f2748030","provider":"Llama 3.3 70B","model":"Llama 3.3 70B","provider_id":"raw","model_id":"meta-llama/llama-3.3-70b-instruct","dataset":"globalopinionqa","framework":"raw","sps":0.762329,"p_dist":0.635373,"p_rank":0.671881,"p_refuse":0.979733,"jsd":0.364627,"tau":0.343763,"n":100,"ci_lower":0.730587,"ci_upper":0.790563,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.626426,"General Attitudes":0.644481,"Health & Science":0.859111,"International Relations & Security":0.640404,"Politics & Governance":0.698489,"Technology & Digital Life":0.852851,"Trust & Wellbeing":0.496796},"topic_metrics":{"Economy & Work":{"sps":0.626426,"n":5,"p_dist":0.661236,"p_rank":0.591615,"p_refuse":0.966928},"General Attitudes":{"sps":0.644481,"n":11,"p_dist":0.665912,"p_rank":0.62305,"p_refuse":1},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.640404,"n":50,"p_dist":0.636381,"p_rank":0.644427,"p_refuse":0.979791},"Politics & Governance":{"sps":0.698489,"n":22,"p_dist":0.642981,"p_rank":0.753997,"p_refuse":0.972106},"Technology & Digital Life":{"sps":0.852851,"n":3,"p_dist":0.802543,"p_rank":0.90316,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.496796,"n":7,"p_dist":0.421563,"p_rank":0.57203,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0.02052,"cost_per_100q":0.02052,"cost_per_sps_point":0.026918,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.654907,"sps_private":0.64909,"sps_public_private_delta":0.005817,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":5,"config_id":"synthpanel--gemini-2.5-flash-lite--tdefault--tplcurrent--ca4dfebb","provider":"SynthPanel (Gemini Flash Lite)","model":"SynthPanel (Gemini Flash Lite)","provider_id":"synthpanel","model_id":"google/gemini-2.5-flash-lite","dataset":"globalopinionqa","framework":"product","sps":0.761601,"p_dist":0.686999,"p_rank":0.624072,"p_refuse":0.973733,"jsd":0.313001,"tau":0.248145,"n":100,"ci_lower":0.72861,"ci_upper":0.792314,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.539409,"General Attitudes":0.682255,"Health & Science":0.654089,"International Relations & Security":0.67247,"Politics & Governance":0.677133,"Technology & Digital Life":0.900638,"Trust & Wellbeing":0.403027},"topic_metrics":{"Economy & Work":{"sps":0.539409,"n":5,"p_dist":0.606354,"p_rank":0.472465,"p_refuse":0.966928},"General Attitudes":{"sps":0.682255,"n":11,"p_dist":0.743676,"p_rank":0.620834,"p_refuse":1},"Health & Science":{"sps":0.654089,"n":2,"p_dist":0.881402,"p_rank":0.426777,"p_refuse":1},"International Relations & Security":{"sps":0.67247,"n":50,"p_dist":0.689814,"p_rank":0.655127,"p_refuse":0.979125},"Politics & Governance":{"sps":0.677133,"n":22,"p_dist":0.724619,"p_rank":0.629647,"p_refuse":0.946349},"Technology & Digital Life":{"sps":0.900638,"n":3,"p_dist":0.849456,"p_rank":0.951821,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.403027,"n":7,"p_dist":0.392035,"p_rank":0.414017,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0.042,"cost_per_100q":0.042,"cost_per_sps_point":0.055147,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.677469,"sps_private":0.577772,"sps_public_private_delta":0.099697,"verification_badge":"flagged","run_count":2,"dataset_coverage_count":3},{"rank":6,"config_id":"raw--gpt-4o-mini--tdefault--tplcurrent--a75c2c4a","provider":"GPT-4o-mini","model":"GPT-4o-mini","provider_id":"raw","model_id":"openai/gpt-4o-mini","dataset":"globalopinionqa","framework":"raw","sps":0.749281,"p_dist":0.633498,"p_rank":0.647946,"p_refuse":0.9664,"jsd":0.366502,"tau":0.295893,"n":100,"ci_lower":0.71347,"ci_upper":0.781585,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.619562,"General Attitudes":0.670304,"Health & Science":0.859111,"International Relations & Security":0.65605,"Politics & Governance":0.629525,"Technology & Digital Life":0.792842,"Trust & Wellbeing":0.407459},"topic_metrics":{"Economy & Work":{"sps":0.619562,"n":5,"p_dist":0.66651,"p_rank":0.572613,"p_refuse":0.966928},"General Attitudes":{"sps":0.670304,"n":11,"p_dist":0.696438,"p_rank":0.644171,"p_refuse":0.951515},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.65605,"n":50,"p_dist":0.655478,"p_rank":0.656623,"p_refuse":0.963791},"Politics & Governance":{"sps":0.629525,"n":22,"p_dist":0.605852,"p_rank":0.653198,"p_refuse":0.972106},"Technology & Digital Life":{"sps":0.792842,"n":3,"p_dist":0.725201,"p_rank":0.860485,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.407459,"n":7,"p_dist":0.356466,"p_rank":0.458453,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0.0315,"cost_per_100q":0.0315,"cost_per_sps_point":0.04204,"cost_per_response":0.000021,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.641713,"sps_private":0.63721,"sps_public_private_delta":0.004503,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":7,"config_id":"raw--claude-sonnet-4.6--tdefault--tplcurrent--94105f12","provider":"Claude Sonnet 4.6","model":"Claude Sonnet 4.6","provider_id":"raw","model_id":"anthropic/claude-sonnet-4.6","dataset":"globalopinionqa","framework":"raw","sps":0.738026,"p_dist":0.592789,"p_rank":0.641557,"p_refuse":0.979733,"jsd":0.407211,"tau":0.283114,"n":100,"ci_lower":0.699732,"ci_upper":0.772631,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.70297,"General Attitudes":0.864428,"Health & Science":0.887379,"International Relations & Security":0.6433,"Politics & Governance":0.514433,"Trust & Wellbeing":0.616127},"topic_metrics":{"Economy & Work":{"sps":0.70297,"n":3,"p_dist":0.769857,"p_rank":0.636083,"p_refuse":0.948618},"General Attitudes":{"sps":0.864428,"n":4,"p_dist":0.785391,"p_rank":0.943465,"p_refuse":1},"Health & Science":{"sps":0.887379,"n":1,"p_dist":0.774759,"p_rank":1,"p_refuse":1},"International Relations & Security":{"sps":0.6433,"n":60,"p_dist":0.618695,"p_rank":0.667906,"p_refuse":0.987269},"Politics & Governance":{"sps":0.514433,"n":30,"p_dist":0.498452,"p_rank":0.530415,"p_refuse":0.963043},"Trust & Wellbeing":{"sps":0.616127,"n":2,"p_dist":0.488861,"p_rank":0.743394,"p_refuse":1}},"demographic_scorecard":null,"cost_usd":0.65418,"cost_per_100q":0.65418,"cost_per_sps_point":0.886392,"cost_per_response":0.000436,"tokens_per_response":128.47,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.625932,"sps_private":0.586117,"sps_public_private_delta":0.039815,"verification_badge":"verified","run_count":1,"dataset_coverage_count":1},{"rank":8,"config_id":"raw--claude-haiku-4-5--tdefault--tplcurrent--662866a7","provider":"Claude Haiku 4.5","model":"Claude Haiku 4.5","provider_id":"raw","model_id":"anthropic/claude-haiku-4-5","dataset":"globalopinionqa","framework":"raw","sps":0.725998,"p_dist":0.600587,"p_rank":0.597675,"p_refuse":0.979733,"jsd":0.399413,"tau":0.195351,"n":100,"ci_lower":0.690318,"ci_upper":0.760315,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.518141,"General Attitudes":0.704727,"Health & Science":0.859111,"International Relations & Security":0.617714,"Politics & Governance":0.558926,"Technology & Digital Life":0.828555,"Trust & Wellbeing":0.312066},"topic_metrics":{"Economy & Work":{"sps":0.518141,"n":5,"p_dist":0.560386,"p_rank":0.475895,"p_refuse":0.966928},"General Attitudes":{"sps":0.704727,"n":11,"p_dist":0.707569,"p_rank":0.701885,"p_refuse":1},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.617714,"n":50,"p_dist":0.634588,"p_rank":0.60084,"p_refuse":0.979791},"Politics & Governance":{"sps":0.558926,"n":22,"p_dist":0.553798,"p_rank":0.564054,"p_refuse":0.972106},"Technology & Digital Life":{"sps":0.828555,"n":3,"p_dist":0.758137,"p_rank":0.898975,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.312066,"n":7,"p_dist":0.243321,"p_rank":0.380811,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0.216,"cost_per_100q":0.216,"cost_per_sps_point":0.297521,"cost_per_response":0.000144,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.613599,"sps_private":0.547836,"sps_public_private_delta":0.065762,"verification_badge":"flagged","run_count":1,"dataset_coverage_count":3},{"rank":9,"config_id":"synthpanel--claude-haiku-4-5--tdefault--tplcurrent--49fbdb2d","provider":"SynthPanel (Haiku 4.5)","model":"SynthPanel (Haiku 4.5)","provider_id":"synthpanel","model_id":"anthropic/claude-haiku-4-5","dataset":"globalopinionqa","framework":"product","sps":0.724855,"p_dist":0.665547,"p_rank":0.628215,"p_refuse":0.880803,"jsd":0.334453,"tau":0.256431,"n":100,"ci_lower":0.685823,"ci_upper":0.760383,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.697658,"General Attitudes":0.65609,"Health & Science":0.859111,"International Relations & Security":0.679484,"Politics & Governance":0.641458,"Technology & Digital Life":0.678827,"Trust & Wellbeing":0.305979},"topic_metrics":{"Economy & Work":{"sps":0.697658,"n":5,"p_dist":0.735985,"p_rank":0.659332,"p_refuse":0.793595},"General Attitudes":{"sps":0.65609,"n":11,"p_dist":0.710557,"p_rank":0.601623,"p_refuse":0.890909},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.679484,"n":50,"p_dist":0.699723,"p_rank":0.659246,"p_refuse":0.908305},"Politics & Governance":{"sps":0.641458,"n":22,"p_dist":0.653976,"p_rank":0.628939,"p_refuse":0.777591},"Technology & Digital Life":{"sps":0.678827,"n":3,"p_dist":0.759139,"p_rank":0.598516,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.305979,"n":7,"p_dist":0.260676,"p_rank":0.351281,"p_refuse":0.98381}},"demographic_scorecard":null,"cost_usd":0.432,"cost_per_100q":0.432,"cost_per_sps_point":0.595981,"cost_per_response":0.000144,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.651432,"sps_private":0.630745,"sps_public_private_delta":0.020688,"verification_badge":"verified","run_count":3,"dataset_coverage_count":3},{"rank":10,"config_id":"baseline--random-baseline--tdefault--tplcurrent--d6b76ef5","provider":"Random Baseline","model":"Random Baseline","provider_id":"baseline","model_id":"random-baseline","dataset":"globalopinionqa","framework":"baseline","sps":0.709683,"p_dist":0.746806,"p_rank":0.399065,"p_refuse":0.983178,"jsd":0.253194,"tau":-0.20187,"n":10,"ci_lower":0.646474,"ci_upper":0.774428,"is_baseline":true,"is_ensemble":false,"samples_per_question":5,"topic_scores":{"General Attitudes":0.435147,"International Relations & Security":0.615386,"Politics & Governance":0.611233,"Trust & Wellbeing":0.379427},"topic_metrics":{"General Attitudes":{"sps":0.435147,"n":1,"p_dist":0.870295,"p_rank":0,"p_refuse":1},"International Relations & Security":{"sps":0.615386,"n":6,"p_dist":0.760529,"p_rank":0.470242,"p_refuse":0.988514},"Politics & Governance":{"sps":0.611233,"n":2,"p_dist":0.776062,"p_rank":0.446403,"p_refuse":0.965346},"Trust & Wellbeing":{"sps":0.379427,"n":1,"p_dist":0.48246,"p_rank":0.276393,"p_refuse":0.97}},"demographic_scorecard":null,"cost_usd":0,"cost_per_100q":0,"cost_per_sps_point":0,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.59385,"sps_private":0.541563,"sps_public_private_delta":0.052287,"verification_badge":"flagged","run_count":1,"dataset_coverage_count":3},{"rank":11,"config_id":"baseline--majority-baseline--tdefault--tplcurrent--e36ac49c","provider":"Majority Baseline","model":"Majority Baseline","provider_id":"baseline","model_id":"majority-baseline","dataset":"globalopinionqa","framework":"baseline","sps":0.68956,"p_dist":0.53424,"p_rank":0.554706,"p_refuse":0.979733,"jsd":0.46576,"tau":0.109412,"n":100,"ci_lower":0.65498,"ci_upper":0.72094,"is_baseline":true,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.32704,"General Attitudes":0.56958,"Health & Science":0.859111,"International Relations & Security":0.585473,"Politics & Governance":0.539684,"Technology & Digital Life":0.591221,"Trust & Wellbeing":0.272589},"topic_metrics":{"Economy & Work":{"sps":0.32704,"n":5,"p_dist":0.345475,"p_rank":0.308605,"p_refuse":0.966928},"General Attitudes":{"sps":0.56958,"n":11,"p_dist":0.628445,"p_rank":0.510714,"p_refuse":1},"Health & Science":{"sps":0.859111,"n":2,"p_dist":0.791446,"p_rank":0.926777,"p_refuse":1},"International Relations & Security":{"sps":0.585473,"n":50,"p_dist":0.575519,"p_rank":0.595428,"p_refuse":0.979791},"Politics & Governance":{"sps":0.539684,"n":22,"p_dist":0.513506,"p_rank":0.565862,"p_refuse":0.972106},"Technology & Digital Life":{"sps":0.591221,"n":3,"p_dist":0.655291,"p_rank":0.527151,"p_refuse":0.967786},"Trust & Wellbeing":{"sps":0.272589,"n":7,"p_dist":0.165987,"p_rank":0.379192,"p_refuse":0.979913}},"demographic_scorecard":null,"cost_usd":0,"cost_per_100q":0,"cost_per_sps_point":0,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.548447,"sps_private":0.530384,"sps_public_private_delta":0.018063,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":1,"config_id":"ensemble--3-model-blend--tdefault--tplcurrent--1bef3e62","provider":"SynthPanel Ensemble (3-model)","model":"SynthPanel Ensemble (3-model)","provider_id":"ensemble","model_id":"3-model-blend","dataset":"opinionsqa","framework":"product","sps":0.876721,"p_dist":0.833241,"p_rank":0.836898,"p_refuse":0.960025,"jsd":0.166759,"tau":0.673797,"n":684,"ci_lower":0.870236,"ci_upper":0.882636,"is_baseline":false,"is_ensemble":true,"topic_scores":{"Economy & Work":0.832038,"General Attitudes":0.844608,"Health & Science":0.851615,"Identity & Demographics":0.809349,"International Relations & Security":0.820213,"Media & Information":0.85288,"Politics & Governance":0.810156,"Social Values & Religion":0.866315,"Technology & Digital Life":0.812016,"Trust & Wellbeing":0.841102},"topic_metrics":{"Economy & Work":{"sps":0.832038,"n":68,"p_dist":0.829105,"p_rank":0.834971,"p_refuse":0.955446},"General Attitudes":{"sps":0.844608,"n":190,"p_dist":0.839781,"p_rank":0.849436,"p_refuse":0.940017},"Health & Science":{"sps":0.851615,"n":47,"p_dist":0.84387,"p_rank":0.85936,"p_refuse":0.975345},"Identity & Demographics":{"sps":0.809349,"n":39,"p_dist":0.82167,"p_rank":0.797028,"p_refuse":0.956395},"International Relations & Security":{"sps":0.820213,"n":149,"p_dist":0.817677,"p_rank":0.822749,"p_refuse":0.972673},"Media & Information":{"sps":0.85288,"n":63,"p_dist":0.856214,"p_rank":0.849547,"p_refuse":0.974872},"Politics & Governance":{"sps":0.810156,"n":40,"p_dist":0.799955,"p_rank":0.820357,"p_refuse":0.960448},"Social Values & Religion":{"sps":0.866315,"n":37,"p_dist":0.860233,"p_rank":0.872398,"p_refuse":0.980768},"Technology & Digital Life":{"sps":0.812016,"n":26,"p_dist":0.821244,"p_rank":0.802787,"p_refuse":0.939822},"Trust & Wellbeing":{"sps":0.841102,"n":25,"p_dist":0.853524,"p_rank":0.828681,"p_refuse":0.978223}},"demographic_scorecard":null,"cost_usd":6.1218,"cost_per_100q":0.895,"cost_per_sps_point":6.982609,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.833043,"sps_private":0.843015,"sps_public_private_delta":0.009972,"verification_badge":"verified","run_count":2,"dataset_coverage_count":3},{"rank":2,"config_id":"raw--gemini-2.5-flash--tdefault--tplcurrent--1579b673","provider":"Gemini 2.5 Flash","model":"Gemini 2.5 Flash","provider_id":"raw","model_id":"google/gemini-2.5-flash","dataset":"opinionsqa","framework":"raw","sps":0.829212,"p_dist":0.737572,"p_rank":0.760539,"p_refuse":0.989527,"jsd":0.262428,"tau":0.521077,"n":684,"ci_lower":0.82167,"ci_upper":0.837119,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.754736,"General Attitudes":0.7406,"Health & Science":0.809566,"Identity & Demographics":0.747952,"International Relations & Security":0.731418,"Media & Information":0.780323,"Politics & Governance":0.719545,"Social Values & Religion":0.757697,"Technology & Digital Life":0.742256,"Trust & Wellbeing":0.753648},"topic_metrics":{"Economy & Work":{"sps":0.754736,"n":68,"p_dist":0.755769,"p_rank":0.753702,"p_refuse":0.990871},"General Attitudes":{"sps":0.7406,"n":190,"p_dist":0.740624,"p_rank":0.740575,"p_refuse":0.989417},"Health & Science":{"sps":0.809566,"n":47,"p_dist":0.771626,"p_rank":0.847506,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.747952,"n":39,"p_dist":0.755022,"p_rank":0.740883,"p_refuse":0.985036},"International Relations & Security":{"sps":0.731418,"n":149,"p_dist":0.69751,"p_rank":0.765325,"p_refuse":0.988117},"Media & Information":{"sps":0.780323,"n":63,"p_dist":0.761743,"p_rank":0.798902,"p_refuse":0.992592},"Politics & Governance":{"sps":0.719545,"n":40,"p_dist":0.741881,"p_rank":0.697208,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.757697,"n":37,"p_dist":0.750134,"p_rank":0.765261,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.742256,"n":26,"p_dist":0.71327,"p_rank":0.771243,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.753648,"n":25,"p_dist":0.751272,"p_rank":0.756024,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":1.61424,"cost_per_100q":0.236,"cost_per_sps_point":1.946716,"cost_per_response":0.000047,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.746696,"sps_private":0.758305,"sps_public_private_delta":0.011609,"verification_badge":"verified","run_count":3,"dataset_coverage_count":3},{"rank":3,"config_id":"synthpanel--claude-sonnet-4--tdefault--tplcurrent--8d1cda33","provider":"SynthPanel (Sonnet 4)","model":"SynthPanel (Sonnet 4)","provider_id":"synthpanel","model_id":"anthropic/claude-sonnet-4","dataset":"opinionsqa","framework":"product","sps":0.828994,"p_dist":0.726347,"p_rank":0.79294,"p_refuse":0.967693,"jsd":0.273653,"tau":0.585881,"n":684,"ci_lower":0.821022,"ci_upper":0.836266,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.746357,"General Attitudes":0.771685,"Health & Science":0.772536,"Identity & Demographics":0.732215,"International Relations & Security":0.735298,"Media & Information":0.779789,"Politics & Governance":0.73616,"Social Values & Religion":0.776574,"Technology & Digital Life":0.826652,"Trust & Wellbeing":0.759988},"topic_metrics":{"Economy & Work":{"sps":0.746357,"n":68,"p_dist":0.709824,"p_rank":0.782889,"p_refuse":0.989466},"General Attitudes":{"sps":0.771685,"n":190,"p_dist":0.741753,"p_rank":0.801618,"p_refuse":0.962927},"Health & Science":{"sps":0.772536,"n":47,"p_dist":0.744551,"p_rank":0.800521,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.732215,"n":39,"p_dist":0.696479,"p_rank":0.767951,"p_refuse":0.983961},"International Relations & Security":{"sps":0.735298,"n":149,"p_dist":0.693615,"p_rank":0.776981,"p_refuse":0.986473},"Media & Information":{"sps":0.779789,"n":63,"p_dist":0.764546,"p_rank":0.795032,"p_refuse":0.892623},"Politics & Governance":{"sps":0.73616,"n":40,"p_dist":0.693288,"p_rank":0.779032,"p_refuse":0.961642},"Social Values & Religion":{"sps":0.776574,"n":37,"p_dist":0.743952,"p_rank":0.809197,"p_refuse":0.989613},"Technology & Digital Life":{"sps":0.826652,"n":26,"p_dist":0.792621,"p_rank":0.860683,"p_refuse":0.91308},"Trust & Wellbeing":{"sps":0.759988,"n":25,"p_dist":0.72332,"p_rank":0.796656,"p_refuse":0.994169}},"demographic_scorecard":null,"cost_usd":8.86464,"cost_per_100q":1.296,"cost_per_sps_point":10.69325,"cost_per_response":0.000432,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.757878,"sps_private":0.766568,"sps_public_private_delta":0.00869,"verification_badge":"verified","run_count":3,"dataset_coverage_count":1},{"rank":4,"config_id":"synthpanel--claude-haiku-4-5--tdefault--tplcurrent--555a9f1e","provider":"SynthPanel (Haiku 4.5)","model":"SynthPanel (Haiku 4.5)","provider_id":"synthpanel","model_id":"anthropic/claude-haiku-4-5","dataset":"opinionsqa","framework":"product","sps":0.828945,"p_dist":0.735935,"p_rank":0.795222,"p_refuse":0.955676,"jsd":0.264065,"tau":0.590444,"n":684,"ci_lower":0.819286,"ci_upper":0.837645,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.75994,"General Attitudes":0.765708,"Health & Science":0.79119,"Identity & Demographics":0.781914,"International Relations & Security":0.734832,"Media & Information":0.785804,"Politics & Governance":0.79692,"Social Values & Religion":0.808175,"Technology & Digital Life":0.78831,"Trust & Wellbeing":0.701753},"topic_metrics":{"Economy & Work":{"sps":0.75994,"n":68,"p_dist":0.725644,"p_rank":0.794237,"p_refuse":0.992994},"General Attitudes":{"sps":0.765708,"n":190,"p_dist":0.736221,"p_rank":0.795194,"p_refuse":0.894496},"Health & Science":{"sps":0.79119,"n":47,"p_dist":0.768191,"p_rank":0.814189,"p_refuse":0.987117},"Identity & Demographics":{"sps":0.781914,"n":39,"p_dist":0.753926,"p_rank":0.809902,"p_refuse":0.984711},"International Relations & Security":{"sps":0.734832,"n":149,"p_dist":0.699251,"p_rank":0.770414,"p_refuse":0.977637},"Media & Information":{"sps":0.785804,"n":63,"p_dist":0.770205,"p_rank":0.801404,"p_refuse":0.937473},"Politics & Governance":{"sps":0.79692,"n":40,"p_dist":0.770551,"p_rank":0.823289,"p_refuse":0.987516},"Social Values & Religion":{"sps":0.808175,"n":37,"p_dist":0.787339,"p_rank":0.82901,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.78831,"n":26,"p_dist":0.761433,"p_rank":0.815188,"p_refuse":0.986501},"Trust & Wellbeing":{"sps":0.701753,"n":25,"p_dist":0.647344,"p_rank":0.756162,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":4.9248,"cost_per_100q":0.72,"cost_per_sps_point":5.941046,"cost_per_response":0.000144,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.763167,"sps_private":0.775035,"sps_public_private_delta":0.011868,"verification_badge":"verified","run_count":15,"dataset_coverage_count":3},{"rank":5,"config_id":"synthpanel--gpt-4o-mini--tdefault--tplcurrent--b1d864c3","provider":"SynthPanel (GPT-4o-mini)","model":"SynthPanel (GPT-4o-mini)","provider_id":"synthpanel","model_id":"openai/gpt-4o-mini","dataset":"opinionsqa","framework":"product","sps":0.822536,"p_dist":0.707743,"p_rank":0.778018,"p_refuse":0.981846,"jsd":0.292257,"tau":0.556037,"n":684,"ci_lower":0.815009,"ci_upper":0.830276,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.752134,"General Attitudes":0.74137,"Health & Science":0.786261,"Identity & Demographics":0.679281,"International Relations & Security":0.722882,"Media & Information":0.790538,"Politics & Governance":0.744189,"Social Values & Religion":0.774435,"Technology & Digital Life":0.683863,"Trust & Wellbeing":0.758541},"topic_metrics":{"Economy & Work":{"sps":0.752134,"n":68,"p_dist":0.723728,"p_rank":0.78054,"p_refuse":0.990871},"General Attitudes":{"sps":0.74137,"n":190,"p_dist":0.716515,"p_rank":0.766225,"p_refuse":0.966233},"Health & Science":{"sps":0.786261,"n":47,"p_dist":0.746984,"p_rank":0.825539,"p_refuse":0.986936},"Identity & Demographics":{"sps":0.679281,"n":39,"p_dist":0.629537,"p_rank":0.729025,"p_refuse":0.985003},"International Relations & Security":{"sps":0.722882,"n":149,"p_dist":0.683135,"p_rank":0.762628,"p_refuse":0.984195},"Media & Information":{"sps":0.790538,"n":63,"p_dist":0.760442,"p_rank":0.820634,"p_refuse":0.992711},"Politics & Governance":{"sps":0.744189,"n":40,"p_dist":0.69424,"p_rank":0.794139,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.774435,"n":37,"p_dist":0.736596,"p_rank":0.812274,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.683863,"n":26,"p_dist":0.633245,"p_rank":0.73448,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.758541,"n":25,"p_dist":0.716081,"p_rank":0.801001,"p_refuse":0.984823}},"demographic_scorecard":null,"cost_usd":0.7182,"cost_per_100q":0.105,"cost_per_sps_point":0.873153,"cost_per_response":0.000021,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.737558,"sps_private":0.76375,"sps_public_private_delta":0.026192,"verification_badge":"verified","run_count":14,"dataset_coverage_count":3},{"rank":6,"config_id":"raw--llama-3.3-70b-instruct--tdefault--tplcurrent--6480fa3c","provider":"Llama 3.3 70B","model":"Llama 3.3 70B","provider_id":"raw","model_id":"meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","framework":"raw","sps":0.81898,"p_dist":0.69294,"p_rank":0.774474,"p_refuse":0.989527,"jsd":0.30706,"tau":0.548948,"n":684,"ci_lower":0.811687,"ci_upper":0.826215,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.739673,"General Attitudes":0.726016,"Health & Science":0.793948,"Identity & Demographics":0.733951,"International Relations & Security":0.722053,"Media & Information":0.74217,"Politics & Governance":0.724096,"Social Values & Religion":0.731249,"Technology & Digital Life":0.712331,"Trust & Wellbeing":0.751661},"topic_metrics":{"Economy & Work":{"sps":0.739673,"n":68,"p_dist":0.704768,"p_rank":0.774579,"p_refuse":0.990871},"General Attitudes":{"sps":0.726016,"n":190,"p_dist":0.704766,"p_rank":0.747267,"p_refuse":0.989417},"Health & Science":{"sps":0.793948,"n":47,"p_dist":0.752668,"p_rank":0.835229,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.733951,"n":39,"p_dist":0.659219,"p_rank":0.808682,"p_refuse":0.985036},"International Relations & Security":{"sps":0.722053,"n":149,"p_dist":0.666607,"p_rank":0.7775,"p_refuse":0.988117},"Media & Information":{"sps":0.74217,"n":63,"p_dist":0.709509,"p_rank":0.774832,"p_refuse":0.992592},"Politics & Governance":{"sps":0.724096,"n":40,"p_dist":0.660775,"p_rank":0.787418,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.731249,"n":37,"p_dist":0.685949,"p_rank":0.77655,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.712331,"n":26,"p_dist":0.681482,"p_rank":0.743181,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.751661,"n":25,"p_dist":0.700116,"p_rank":0.803206,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":0.467856,"cost_per_100q":0.0684,"cost_per_sps_point":0.571267,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.729757,"sps_private":0.749195,"sps_public_private_delta":0.019438,"verification_badge":"verified","run_count":9,"dataset_coverage_count":3},{"rank":7,"config_id":"synthpanel--gemini-2.5-flash-lite--tdefault--tplcurrent--5a08bef0","provider":"SynthPanel (Gemini Flash Lite)","model":"SynthPanel (Gemini Flash Lite)","provider_id":"synthpanel","model_id":"google/gemini-2.5-flash-lite","dataset":"opinionsqa","framework":"product","sps":0.81604,"p_dist":0.748961,"p_rank":0.765861,"p_refuse":0.933296,"jsd":0.251039,"tau":0.531723,"n":684,"ci_lower":0.80675,"ci_upper":0.825453,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.75061,"General Attitudes":0.774235,"Health & Science":0.79697,"Identity & Demographics":0.684477,"International Relations & Security":0.760496,"Media & Information":0.7774,"Politics & Governance":0.680481,"Social Values & Religion":0.773925,"Technology & Digital Life":0.730442,"Trust & Wellbeing":0.745398},"topic_metrics":{"Economy & Work":{"sps":0.75061,"n":68,"p_dist":0.741964,"p_rank":0.759256,"p_refuse":0.86889},"General Attitudes":{"sps":0.774235,"n":190,"p_dist":0.773262,"p_rank":0.775207,"p_refuse":0.947377},"Health & Science":{"sps":0.79697,"n":47,"p_dist":0.788261,"p_rank":0.805679,"p_refuse":0.945157},"Identity & Demographics":{"sps":0.684477,"n":39,"p_dist":0.700572,"p_rank":0.668383,"p_refuse":0.880343},"International Relations & Security":{"sps":0.760496,"n":149,"p_dist":0.744704,"p_rank":0.776288,"p_refuse":0.950112},"Media & Information":{"sps":0.7774,"n":63,"p_dist":0.766087,"p_rank":0.788713,"p_refuse":0.988785},"Politics & Governance":{"sps":0.680481,"n":40,"p_dist":0.648088,"p_rank":0.712873,"p_refuse":0.891615},"Social Values & Religion":{"sps":0.773925,"n":37,"p_dist":0.763666,"p_rank":0.784185,"p_refuse":0.959146},"Technology & Digital Life":{"sps":0.730442,"n":26,"p_dist":0.697065,"p_rank":0.763818,"p_refuse":0.834881},"Trust & Wellbeing":{"sps":0.745398,"n":25,"p_dist":0.760725,"p_rank":0.730071,"p_refuse":0.952505}},"demographic_scorecard":null,"cost_usd":0.4788,"cost_per_100q":0.07,"cost_per_sps_point":0.586736,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.755437,"sps_private":0.76515,"sps_public_private_delta":0.009713,"verification_badge":"verified","run_count":6,"dataset_coverage_count":3},{"rank":8,"config_id":"raw--claude-haiku-4-5--tdefault--tplcurrent--9bd14231","provider":"Claude Haiku 4.5","model":"Claude Haiku 4.5","provider_id":"raw","model_id":"anthropic/claude-haiku-4-5","dataset":"opinionsqa","framework":"raw","sps":0.815296,"p_dist":0.689574,"p_rank":0.766787,"p_refuse":0.989527,"jsd":0.310426,"tau":0.533574,"n":684,"ci_lower":0.80748,"ci_upper":0.822979,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.734091,"General Attitudes":0.724052,"Health & Science":0.780889,"Identity & Demographics":0.68391,"International Relations & Security":0.725823,"Media & Information":0.765082,"Politics & Governance":0.668284,"Social Values & Religion":0.731056,"Technology & Digital Life":0.721434,"Trust & Wellbeing":0.733109},"topic_metrics":{"Economy & Work":{"sps":0.734091,"n":68,"p_dist":0.700412,"p_rank":0.76777,"p_refuse":0.990871},"General Attitudes":{"sps":0.724052,"n":190,"p_dist":0.689532,"p_rank":0.758573,"p_refuse":0.989417},"Health & Science":{"sps":0.780889,"n":47,"p_dist":0.740982,"p_rank":0.820796,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.68391,"n":39,"p_dist":0.647543,"p_rank":0.720276,"p_refuse":0.985036},"International Relations & Security":{"sps":0.725823,"n":149,"p_dist":0.676542,"p_rank":0.775103,"p_refuse":0.988117},"Media & Information":{"sps":0.765082,"n":63,"p_dist":0.739328,"p_rank":0.790835,"p_refuse":0.992592},"Politics & Governance":{"sps":0.668284,"n":40,"p_dist":0.640339,"p_rank":0.69623,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.731056,"n":37,"p_dist":0.681507,"p_rank":0.780604,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.721434,"n":26,"p_dist":0.673774,"p_rank":0.769094,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.733109,"n":25,"p_dist":0.688773,"p_rank":0.777444,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":4.9248,"cost_per_100q":0.72,"cost_per_sps_point":6.040506,"cost_per_response":0.000144,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.722774,"sps_private":0.74938,"sps_public_private_delta":0.026606,"verification_badge":"verified","run_count":28,"dataset_coverage_count":3},{"rank":9,"config_id":"raw--gpt-4o--tdefault--tplcurrent--ec90182d","provider":"GPT-4o","model":"GPT-4o","provider_id":"raw","model_id":"openai/gpt-4o","dataset":"opinionsqa","framework":"raw","sps":0.813134,"p_dist":0.726307,"p_rank":0.720431,"p_refuse":0.992662,"jsd":0.273693,"tau":0.440863,"n":200,"ci_lower":0.796357,"ci_upper":0.828225,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.73882,"General Attitudes":0.717188,"Health & Science":0.828446,"Identity & Demographics":0.672419,"International Relations & Security":0.684696,"Media & Information":0.700956,"Politics & Governance":0.824484,"Social Values & Religion":0.690722,"Technology & Digital Life":0.792024,"Trust & Wellbeing":0.66958},"topic_metrics":{"Economy & Work":{"sps":0.73882,"n":27,"p_dist":0.724089,"p_rank":0.753552,"p_refuse":0.994308},"General Attitudes":{"sps":0.717188,"n":79,"p_dist":0.736504,"p_rank":0.697873,"p_refuse":0.991424},"Health & Science":{"sps":0.828446,"n":19,"p_dist":0.818353,"p_rank":0.83854,"p_refuse":0.994827},"Identity & Demographics":{"sps":0.672419,"n":5,"p_dist":0.55749,"p_rank":0.787347,"p_refuse":0.985086},"International Relations & Security":{"sps":0.684696,"n":38,"p_dist":0.680114,"p_rank":0.689278,"p_refuse":0.993962},"Media & Information":{"sps":0.700956,"n":2,"p_dist":0.788699,"p_rank":0.613214,"p_refuse":0.996077},"Politics & Governance":{"sps":0.824484,"n":3,"p_dist":0.780769,"p_rank":0.8682,"p_refuse":0.983405},"Social Values & Religion":{"sps":0.690722,"n":18,"p_dist":0.715551,"p_rank":0.665894,"p_refuse":0.991819},"Technology & Digital Life":{"sps":0.792024,"n":5,"p_dist":0.775129,"p_rank":0.80892,"p_refuse":0.995632},"Trust & Wellbeing":{"sps":0.66958,"n":4,"p_dist":0.66787,"p_rank":0.67129,"p_refuse":0.99816}},"demographic_scorecard":null,"cost_usd":3.5,"cost_per_100q":1.75,"cost_per_sps_point":4.304334,"cost_per_response":0.00035,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.719324,"sps_private":0.736913,"sps_public_private_delta":0.01759,"verification_badge":"verified","run_count":2,"dataset_coverage_count":1},{"rank":10,"config_id":"raw--gpt-4o-mini--tdefault--tplcurrent--ede40dcf","provider":"GPT-4o-mini","model":"GPT-4o-mini","provider_id":"raw","model_id":"openai/gpt-4o-mini","dataset":"opinionsqa","framework":"raw","sps":0.812518,"p_dist":0.686116,"p_rank":0.76191,"p_refuse":0.989527,"jsd":0.313884,"tau":0.52382,"n":684,"ci_lower":0.80487,"ci_upper":0.820012,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.737991,"General Attitudes":0.708131,"Health & Science":0.777204,"Identity & Demographics":0.711758,"International Relations & Security":0.703127,"Media & Information":0.759826,"Politics & Governance":0.751486,"Social Values & Religion":0.712703,"Technology & Digital Life":0.705445,"Trust & Wellbeing":0.752137},"topic_metrics":{"Economy & Work":{"sps":0.737991,"n":68,"p_dist":0.700631,"p_rank":0.77535,"p_refuse":0.990871},"General Attitudes":{"sps":0.708131,"n":190,"p_dist":0.685296,"p_rank":0.730967,"p_refuse":0.989417},"Health & Science":{"sps":0.777204,"n":47,"p_dist":0.731353,"p_rank":0.823054,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.711758,"n":39,"p_dist":0.669219,"p_rank":0.754298,"p_refuse":0.985036},"International Relations & Security":{"sps":0.703127,"n":149,"p_dist":0.654969,"p_rank":0.751285,"p_refuse":0.988117},"Media & Information":{"sps":0.759826,"n":63,"p_dist":0.728107,"p_rank":0.791546,"p_refuse":0.992592},"Politics & Governance":{"sps":0.751486,"n":40,"p_dist":0.693591,"p_rank":0.809381,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.712703,"n":37,"p_dist":0.662307,"p_rank":0.763098,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.705445,"n":26,"p_dist":0.658957,"p_rank":0.751933,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.752137,"n":25,"p_dist":0.725532,"p_rank":0.778743,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":0.7182,"cost_per_100q":0.105,"cost_per_sps_point":0.883919,"cost_per_response":0.000021,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.7228,"sps_private":0.728771,"sps_public_private_delta":0.005971,"verification_badge":"verified","run_count":18,"dataset_coverage_count":3},{"rank":11,"config_id":"raw--claude-sonnet-4--tdefault--tplcurrent--482c42c7","provider":"Claude Sonnet 4","model":"Claude Sonnet 4","provider_id":"raw","model_id":"anthropic/claude-sonnet-4","dataset":"opinionsqa","framework":"raw","sps":0.782456,"p_dist":0.648247,"p_rank":0.709592,"p_refuse":0.989527,"jsd":0.351753,"tau":0.419185,"n":684,"ci_lower":0.772284,"ci_upper":0.792899,"is_baseline":false,"is_ensemble":false,"samples_per_question":50,"topic_scores":{"Economy & Work":0.616553,"General Attitudes":0.689824,"Health & Science":0.711565,"Identity & Demographics":0.641569,"International Relations & Security":0.688858,"Media & Information":0.777605,"Politics & Governance":0.626659,"Social Values & Religion":0.569164,"Technology & Digital Life":0.728625,"Trust & Wellbeing":0.649029},"topic_metrics":{"Economy & Work":{"sps":0.616553,"n":68,"p_dist":0.589159,"p_rank":0.643948,"p_refuse":0.990871},"General Attitudes":{"sps":0.689824,"n":190,"p_dist":0.663699,"p_rank":0.715949,"p_refuse":0.989417},"Health & Science":{"sps":0.711565,"n":47,"p_dist":0.688065,"p_rank":0.735065,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.641569,"n":39,"p_dist":0.619953,"p_rank":0.663185,"p_refuse":0.985036},"International Relations & Security":{"sps":0.688858,"n":149,"p_dist":0.64487,"p_rank":0.732847,"p_refuse":0.988117},"Media & Information":{"sps":0.777605,"n":63,"p_dist":0.760349,"p_rank":0.794861,"p_refuse":0.992592},"Politics & Governance":{"sps":0.626659,"n":40,"p_dist":0.58007,"p_rank":0.673248,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.569164,"n":37,"p_dist":0.536135,"p_rank":0.602192,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.728625,"n":26,"p_dist":0.693701,"p_rank":0.763548,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.649029,"n":25,"p_dist":0.626195,"p_rank":0.671864,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":14.7744,"cost_per_100q":2.16,"cost_per_sps_point":18.882084,"cost_per_response":0.000432,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.6762,"sps_private":0.689586,"sps_public_private_delta":0.013386,"verification_badge":"verified","run_count":3,"dataset_coverage_count":1},{"rank":12,"config_id":"baseline--random-baseline--tdefault--tplcurrent--971f8ad1","provider":"Random Baseline","model":"Random Baseline","provider_id":"baseline","model_id":"random-baseline","dataset":"opinionsqa","framework":"baseline","sps":0.762852,"p_dist":0.805999,"p_rank":0.49303,"p_refuse":0.989527,"jsd":0.194001,"tau":-0.01394,"n":684,"ci_lower":0.755076,"ci_upper":0.770623,"is_baseline":true,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.645759,"General Attitudes":0.635076,"Health & Science":0.667831,"Identity & Demographics":0.625835,"International Relations & Security":0.659696,"Media & Information":0.671985,"Politics & Governance":0.658529,"Social Values & Religion":0.650374,"Technology & Digital Life":0.660403,"Trust & Wellbeing":0.627637},"topic_metrics":{"Economy & Work":{"sps":0.645759,"n":68,"p_dist":0.795882,"p_rank":0.495635,"p_refuse":0.990871},"General Attitudes":{"sps":0.635076,"n":190,"p_dist":0.810341,"p_rank":0.459811,"p_refuse":0.989417},"Health & Science":{"sps":0.667831,"n":47,"p_dist":0.803667,"p_rank":0.531994,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.625835,"n":39,"p_dist":0.80696,"p_rank":0.444709,"p_refuse":0.985036},"International Relations & Security":{"sps":0.659696,"n":149,"p_dist":0.816773,"p_rank":0.502619,"p_refuse":0.988117},"Media & Information":{"sps":0.671985,"n":63,"p_dist":0.800015,"p_rank":0.543956,"p_refuse":0.992592},"Politics & Governance":{"sps":0.658529,"n":40,"p_dist":0.813828,"p_rank":0.503231,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.650374,"n":37,"p_dist":0.79538,"p_rank":0.505368,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.660403,"n":26,"p_dist":0.785307,"p_rank":0.535498,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.627637,"n":25,"p_dist":0.778964,"p_rank":0.47631,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":0,"cost_per_100q":0,"cost_per_sps_point":0,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.65348,"sps_private":0.633965,"sps_public_private_delta":0.019515,"verification_badge":"verified","run_count":15,"dataset_coverage_count":3},{"rank":13,"config_id":"raw--gemini-2.5-flash-lite--tdefault--tplcurrent--bdb40b06","provider":"Gemini Flash Lite","model":"Gemini Flash Lite","provider_id":"raw","model_id":"google/gemini-2.5-flash-lite","dataset":"opinionsqa","framework":"raw","sps":0.737838,"p_dist":0.734729,"p_rank":0.740947,"p_refuse":0,"jsd":0.265271,"tau":0.481894,"n":100,"ci_lower":0.70442,"ci_upper":0.769157,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.765408,"General Attitudes":0.736044,"Health & Science":0.768253,"International Relations & Security":0.659767,"Media & Information":0.700836,"Politics & Governance":0.763278,"Social Values & Religion":0.798768,"Technology & Digital Life":0.763929},"topic_metrics":{"Economy & Work":{"sps":0.765408,"n":20,"p_dist":0.749863,"p_rank":0.780952},"General Attitudes":{"sps":0.736044,"n":38,"p_dist":0.739251,"p_rank":0.732836},"Health & Science":{"sps":0.768253,"n":1,"p_dist":0.762644,"p_rank":0.773861},"International Relations & Security":{"sps":0.659767,"n":19,"p_dist":0.660753,"p_rank":0.658781},"Media & Information":{"sps":0.700836,"n":2,"p_dist":0.743559,"p_rank":0.658114},"Politics & Governance":{"sps":0.763278,"n":1,"p_dist":0.727749,"p_rank":0.798807},"Social Values & Religion":{"sps":0.798768,"n":15,"p_dist":0.793966,"p_rank":0.80357},"Technology & Digital Life":{"sps":0.763929,"n":4,"p_dist":0.735699,"p_rank":0.792159}},"demographic_scorecard":null,"cost_usd":0.042,"cost_per_100q":0.042,"cost_per_sps_point":0.056923,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.736515,"sps_private":0.742269,"sps_public_private_delta":0.005755,"verification_badge":"verified","run_count":2,"dataset_coverage_count":1},{"rank":14,"config_id":"baseline--majority-baseline--tdefault--tplcurrent--04e61fe9","provider":"Majority Baseline","model":"Majority Baseline","provider_id":"baseline","model_id":"majority-baseline","dataset":"opinionsqa","framework":"baseline","sps":0.704504,"p_dist":0.508148,"p_rank":0.615837,"p_refuse":0.989527,"jsd":0.491852,"tau":0.231674,"n":684,"ci_lower":0.693979,"ci_upper":0.714322,"is_baseline":true,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.541087,"General Attitudes":0.578304,"Health & Science":0.651608,"Identity & Demographics":0.582434,"International Relations & Security":0.521504,"Media & Information":0.628286,"Politics & Governance":0.538986,"Social Values & Religion":0.547032,"Technology & Digital Life":0.471897,"Trust & Wellbeing":0.521415},"topic_metrics":{"Economy & Work":{"sps":0.541087,"n":68,"p_dist":0.492042,"p_rank":0.590133,"p_refuse":0.990871},"General Attitudes":{"sps":0.578304,"n":190,"p_dist":0.530955,"p_rank":0.625653,"p_refuse":0.989417},"Health & Science":{"sps":0.651608,"n":47,"p_dist":0.60807,"p_rank":0.695146,"p_refuse":0.987072},"Identity & Demographics":{"sps":0.582434,"n":39,"p_dist":0.526538,"p_rank":0.638331,"p_refuse":0.985036},"International Relations & Security":{"sps":0.521504,"n":149,"p_dist":0.465002,"p_rank":0.578007,"p_refuse":0.988117},"Media & Information":{"sps":0.628286,"n":63,"p_dist":0.582391,"p_rank":0.67418,"p_refuse":0.992592},"Politics & Governance":{"sps":0.538986,"n":40,"p_dist":0.46337,"p_rank":0.614602,"p_refuse":0.988829},"Social Values & Religion":{"sps":0.547032,"n":37,"p_dist":0.483768,"p_rank":0.610297,"p_refuse":0.990371},"Technology & Digital Life":{"sps":0.471897,"n":26,"p_dist":0.393125,"p_rank":0.55067,"p_refuse":0.992887},"Trust & Wellbeing":{"sps":0.521415,"n":25,"p_dist":0.459484,"p_rank":0.583346,"p_refuse":0.995372}},"demographic_scorecard":null,"cost_usd":0,"cost_per_100q":0,"cost_per_sps_point":0,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.562208,"sps_private":0.561147,"sps_public_private_delta":0.001061,"verification_badge":"verified","run_count":15,"dataset_coverage_count":3},{"rank":1,"config_id":"ensemble--3-model-blend--tdefault--tplcurrent--ef75a647","provider":"SynthPanel Ensemble (3-model)","model":"SynthPanel Ensemble (3-model)","provider_id":"ensemble","model_id":"3-model-blend","dataset":"subpop","framework":"product","sps":0.830664,"p_dist":0.791388,"p_rank":0.805484,"p_refuse":0.895119,"jsd":0.208612,"tau":0.610968,"n":200,"ci_lower":0.816957,"ci_upper":0.843289,"is_baseline":false,"is_ensemble":true,"topic_scores":{"Economy & Work":0.762373,"General Attitudes":0.845385,"Health & Science":0.812754,"Identity & Demographics":0.7932,"International Relations & Security":0.814971,"Media & Information":0.756949,"Politics & Governance":0.790507,"Social Values & Religion":0.795281,"Technology & Digital Life":0.764256,"Trust & Wellbeing":0.872348},"topic_metrics":{"Economy & Work":{"sps":0.762373,"n":26,"p_dist":0.742429,"p_rank":0.782317,"p_refuse":0.923305},"General Attitudes":{"sps":0.845385,"n":33,"p_dist":0.851675,"p_rank":0.839095,"p_refuse":0.811145},"Health & Science":{"sps":0.812754,"n":6,"p_dist":0.872164,"p_rank":0.753344,"p_refuse":0.885327},"Identity & Demographics":{"sps":0.7932,"n":1,"p_dist":0.812539,"p_rank":0.773861,"p_refuse":0.916253},"International Relations & Security":{"sps":0.814971,"n":38,"p_dist":0.798086,"p_rank":0.831857,"p_refuse":0.912695},"Media & Information":{"sps":0.756949,"n":1,"p_dist":0.806877,"p_rank":0.70702,"p_refuse":0.900234},"Politics & Governance":{"sps":0.790507,"n":23,"p_dist":0.79085,"p_rank":0.790164,"p_refuse":0.911156},"Social Values & Religion":{"sps":0.795281,"n":37,"p_dist":0.77816,"p_rank":0.812403,"p_refuse":0.899625},"Technology & Digital Life":{"sps":0.764256,"n":33,"p_dist":0.754101,"p_rank":0.774411,"p_refuse":0.928111},"Trust & Wellbeing":{"sps":0.872348,"n":2,"p_dist":0.911361,"p_rank":0.833333,"p_refuse":0.784447}},"demographic_scorecard":null,"cost_usd":1.074,"cost_per_100q":0.537,"cost_per_sps_point":1.292942,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.794317,"sps_private":0.809572,"sps_public_private_delta":0.015254,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":2,"config_id":"synthpanel--gemini-2.5-flash-lite--tdefault--tplcurrent--650ec804","provider":"SynthPanel (Gemini Flash Lite)","model":"SynthPanel (Gemini Flash Lite)","provider_id":"synthpanel","model_id":"google/gemini-2.5-flash-lite","dataset":"subpop","framework":"product","sps":0.820993,"p_dist":0.70706,"p_rank":0.780386,"p_refuse":0.975532,"jsd":0.29294,"tau":0.560772,"n":200,"ci_lower":0.806912,"ci_upper":0.833803,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.751091,"General Attitudes":0.7825,"Health & Science":0.828025,"Identity & Demographics":0.845662,"International Relations & Security":0.740791,"Politics & Governance":0.701071,"Social Values & Religion":0.749675,"Technology & Digital Life":0.709148,"Trust & Wellbeing":0.924913},"topic_metrics":{"Economy & Work":{"sps":0.751091,"n":17,"p_dist":0.7144,"p_rank":0.787781,"p_refuse":0.991415},"General Attitudes":{"sps":0.7825,"n":37,"p_dist":0.753569,"p_rank":0.811432,"p_refuse":0.913336},"Health & Science":{"sps":0.828025,"n":5,"p_dist":0.859856,"p_rank":0.796194,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.845662,"n":1,"p_dist":0.837771,"p_rank":0.853554,"p_refuse":0.994858},"International Relations & Security":{"sps":0.740791,"n":33,"p_dist":0.689805,"p_rank":0.791777,"p_refuse":0.991075},"Politics & Governance":{"sps":0.701071,"n":22,"p_dist":0.662488,"p_rank":0.739654,"p_refuse":0.983152},"Social Values & Religion":{"sps":0.749675,"n":36,"p_dist":0.69695,"p_rank":0.802401,"p_refuse":0.986631},"Technology & Digital Life":{"sps":0.709148,"n":47,"p_dist":0.681452,"p_rank":0.736844,"p_refuse":0.992881},"Trust & Wellbeing":{"sps":0.924913,"n":2,"p_dist":0.895701,"p_rank":0.954124,"p_refuse":0.987774}},"demographic_scorecard":null,"cost_usd":0.084,"cost_per_100q":0.042,"cost_per_sps_point":0.102315,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.742098,"sps_private":0.748118,"sps_public_private_delta":0.00602,"verification_badge":"verified","run_count":2,"dataset_coverage_count":3},{"rank":3,"config_id":"synthpanel--claude-haiku-4-5--tdefault--tplcurrent--fe74b735","provider":"SynthPanel (Haiku 4.5)","model":"SynthPanel (Haiku 4.5)","provider_id":"synthpanel","model_id":"anthropic/claude-haiku-4-5","dataset":"subpop","framework":"product","sps":0.809082,"p_dist":0.712052,"p_rank":0.757229,"p_refuse":0.957965,"jsd":0.287948,"tau":0.514458,"n":200,"ci_lower":0.79197,"ci_upper":0.824486,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.706841,"General Attitudes":0.768096,"Health & Science":0.771141,"Identity & Demographics":0.534312,"International Relations & Security":0.764537,"Politics & Governance":0.683359,"Social Values & Religion":0.726185,"Technology & Digital Life":0.728579,"Trust & Wellbeing":0.726388},"topic_metrics":{"Economy & Work":{"sps":0.706841,"n":17,"p_dist":0.704346,"p_rank":0.709335,"p_refuse":0.991415},"General Attitudes":{"sps":0.768096,"n":37,"p_dist":0.761268,"p_rank":0.774923,"p_refuse":0.868046},"Health & Science":{"sps":0.771141,"n":5,"p_dist":0.84695,"p_rank":0.695331,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.534312,"n":1,"p_dist":0.477337,"p_rank":0.591287,"p_refuse":0.994858},"International Relations & Security":{"sps":0.764537,"n":33,"p_dist":0.739646,"p_rank":0.789427,"p_refuse":0.99147},"Politics & Governance":{"sps":0.683359,"n":22,"p_dist":0.62376,"p_rank":0.742959,"p_refuse":0.904004},"Social Values & Religion":{"sps":0.726185,"n":36,"p_dist":0.698913,"p_rank":0.753457,"p_refuse":0.984299},"Technology & Digital Life":{"sps":0.728579,"n":47,"p_dist":0.700745,"p_rank":0.756413,"p_refuse":0.992337},"Trust & Wellbeing":{"sps":0.726388,"n":2,"p_dist":0.66532,"p_rank":0.787458,"p_refuse":0.987774}},"demographic_scores":[{"attribute":"POLPARTY","group":"Democrat","p_dist":0.645286,"p_cond":0.033446,"n_questions":100},{"attribute":"POLPARTY","group":"Republican","p_dist":0.660841,"p_cond":0.071917,"n_questions":100},{"attribute":"INCOME","group":"$100,000 or more","p_dist":0.671439,"p_cond":0.034041,"n_questions":100},{"attribute":"INCOME","group":"Less than $30,000","p_dist":0.60553,"p_cond":0.023,"n_questions":100},{"attribute":"EDUCATION","group":"College graduate/some postgrad","p_dist":0.640846,"p_cond":0.035699,"n_questions":100},{"attribute":"EDUCATION","group":"Less than high school","p_dist":0.596863,"p_cond":0.037923,"n_questions":100},{"attribute":"CREGION","group":"Northeast","p_dist":0.622784,"p_cond":0.018402,"n_questions":100},{"attribute":"CREGION","group":"South","p_dist":0.640679,"p_cond":0.035833,"n_questions":100},{"attribute":"RACE","group":"Asian","p_dist":0.642152,"p_cond":0.029038,"n_questions":100},{"attribute":"RACE","group":"Black","p_dist":0.583925,"p_cond":0.027622,"n_questions":100},{"attribute":"RACE","group":"Hispanic","p_dist":0.609013,"p_cond":0.039195,"n_questions":100},{"attribute":"RACE","group":"White","p_dist":0.792426,"p_cond":0.141474,"n_questions":100},{"attribute":"RELIG","group":"Atheist","p_dist":0.73835,"p_cond":0.09528,"n_questions":20},{"attribute":"RELIG","group":"Hindu","p_dist":0.806252,"p_cond":0.129162,"n_questions":20},{"attribute":"RELIG","group":"Jewish","p_dist":0.773676,"p_cond":0.126971,"n_questions":20},{"attribute":"RELIG","group":"Muslim","p_dist":0.819781,"p_cond":0.161155,"n_questions":20},{"attribute":"RELIG","group":"Protestant","p_dist":0.838025,"p_cond":0.178139,"n_questions":20},{"attribute":"SEX","group":"Female","p_dist":0.617469,"p_cond":0.026534,"n_questions":100},{"attribute":"SEX","group":"Male","p_dist":0.647331,"p_cond":0.054629,"n_questions":100}],"demographic_scorecard":{"dataset":"subpop","dimensions":[{"attribute":"CREGION","label":"Geography (US Census region)","groups":[{"group":"Northeast","score":0.622784,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.018402},{"group":"South","score":0.640679,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.035833}]},{"attribute":"EDUCATION","label":"Education","groups":[{"group":"College graduate/some postgrad","score":0.640846,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.035699},{"group":"Less than high school","score":0.596863,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.037923}]},{"attribute":"INCOME","label":"Income","groups":[{"group":"$100,000 or more","score":0.671439,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.034041},{"group":"Less than $30,000","score":0.60553,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.023}]},{"attribute":"POLPARTY","label":"Political party","groups":[{"group":"Democrat","score":0.645286,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.033446},{"group":"Republican","score":0.660841,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.071917}]},{"attribute":"RACE","label":"Race / ethnicity","groups":[{"group":"Asian","score":0.642152,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.029038},{"group":"Black","score":0.583925,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.027622},{"group":"Hispanic","score":0.609013,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.039195},{"group":"White","score":0.792426,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.141474}]},{"attribute":"RELIG","label":"Religion","groups":[{"group":"Atheist","score":0.73835,"ci_lower":null,"ci_upper":null,"n":20,"p_cond":0.09528},{"group":"Hindu","score":0.806252,"ci_lower":null,"ci_upper":null,"n":20,"p_cond":0.129162},{"group":"Jewish","score":0.773676,"ci_lower":null,"ci_upper":null,"n":20,"p_cond":0.126971},{"group":"Muslim","score":0.819781,"ci_lower":null,"ci_upper":null,"n":20,"p_cond":0.161155},{"group":"Protestant","score":0.838025,"ci_lower":null,"ci_upper":null,"n":20,"p_cond":0.178139}]},{"attribute":"SEX","label":"Sex","groups":[{"group":"Female","score":0.617469,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.026534},{"group":"Male","score":0.647331,"ci_lower":null,"ci_upper":null,"n":100,"p_cond":0.054629}]}]},"cost_usd":0.864,"cost_per_100q":0.432,"cost_per_sps_point":1.067877,"cost_per_response":0.000144,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.724199,"sps_private":0.76287,"sps_public_private_delta":0.038671,"verification_badge":"verified","run_count":4,"dataset_coverage_count":3},{"rank":4,"config_id":"raw--llama-3.3-70b-instruct--tdefault--tplcurrent--670d96d8","provider":"Llama 3.3 70B","model":"Llama 3.3 70B","provider_id":"raw","model_id":"meta-llama/llama-3.3-70b-instruct","dataset":"subpop","framework":"raw","sps":0.795507,"p_dist":0.6551,"p_rank":0.755888,"p_refuse":0.975532,"jsd":0.3449,"tau":0.511777,"n":200,"ci_lower":0.778252,"ci_upper":0.810479,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.717211,"General Attitudes":0.70399,"Health & Science":0.806688,"Identity & Demographics":0.630652,"International Relations & Security":0.711454,"Politics & Governance":0.703718,"Social Values & Religion":0.715449,"Technology & Digital Life":0.675084,"Trust & Wellbeing":0.874807},"topic_metrics":{"Economy & Work":{"sps":0.717211,"n":17,"p_dist":0.658211,"p_rank":0.77621,"p_refuse":0.991415},"General Attitudes":{"sps":0.70399,"n":37,"p_dist":0.66723,"p_rank":0.74075,"p_refuse":0.913336},"Health & Science":{"sps":0.806688,"n":5,"p_dist":0.813139,"p_rank":0.800238,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.630652,"n":1,"p_dist":0.487443,"p_rank":0.773861,"p_refuse":0.994858},"International Relations & Security":{"sps":0.711454,"n":33,"p_dist":0.662207,"p_rank":0.760701,"p_refuse":0.991075},"Politics & Governance":{"sps":0.703718,"n":22,"p_dist":0.644463,"p_rank":0.762973,"p_refuse":0.983152},"Social Values & Religion":{"sps":0.715449,"n":36,"p_dist":0.648003,"p_rank":0.782896,"p_refuse":0.986631},"Technology & Digital Life":{"sps":0.675084,"n":47,"p_dist":0.630629,"p_rank":0.719538,"p_refuse":0.992881},"Trust & Wellbeing":{"sps":0.874807,"n":2,"p_dist":0.79549,"p_rank":0.954124,"p_refuse":0.987774}},"demographic_scorecard":null,"cost_usd":0.08208,"cost_per_100q":0.04104,"cost_per_sps_point":0.103179,"cost_per_response":0.000014,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.705237,"sps_private":0.706189,"sps_public_private_delta":0.000952,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":5,"config_id":"synthpanel--gpt-4o-mini--tdefault--tplcurrent--b631856b","provider":"SynthPanel (GPT-4o-mini)","model":"SynthPanel (GPT-4o-mini)","provider_id":"synthpanel","model_id":"openai/gpt-4o-mini","dataset":"subpop","framework":"product","sps":0.786779,"p_dist":0.65172,"p_rank":0.733086,"p_refuse":0.975532,"jsd":0.34828,"tau":0.466172,"n":200,"ci_lower":0.771457,"ci_upper":0.802165,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.653807,"General Attitudes":0.71223,"Health & Science":0.687577,"Identity & Demographics":0.505814,"International Relations & Security":0.732399,"Politics & Governance":0.717032,"Social Values & Religion":0.725218,"Technology & Digital Life":0.625384,"Trust & Wellbeing":0.812457},"topic_metrics":{"Economy & Work":{"sps":0.653807,"n":17,"p_dist":0.601009,"p_rank":0.706604,"p_refuse":0.991415},"General Attitudes":{"sps":0.71223,"n":37,"p_dist":0.688489,"p_rank":0.735971,"p_refuse":0.913336},"Health & Science":{"sps":0.687577,"n":5,"p_dist":0.701074,"p_rank":0.67408,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.505814,"n":1,"p_dist":0.393776,"p_rank":0.617851,"p_refuse":0.994858},"International Relations & Security":{"sps":0.732399,"n":33,"p_dist":0.683264,"p_rank":0.781534,"p_refuse":0.991075},"Politics & Governance":{"sps":0.717032,"n":22,"p_dist":0.652762,"p_rank":0.781303,"p_refuse":0.983152},"Social Values & Religion":{"sps":0.725218,"n":36,"p_dist":0.680113,"p_rank":0.770323,"p_refuse":0.986631},"Technology & Digital Life":{"sps":0.625384,"n":47,"p_dist":0.59102,"p_rank":0.659747,"p_refuse":0.992881},"Trust & Wellbeing":{"sps":0.812457,"n":2,"p_dist":0.791581,"p_rank":0.833333,"p_refuse":0.987774}},"demographic_scorecard":null,"cost_usd":0.126,"cost_per_100q":0.063,"cost_per_sps_point":0.160147,"cost_per_response":0.000021,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.692576,"sps_private":0.691936,"sps_public_private_delta":0.00064,"verification_badge":"verified","run_count":3,"dataset_coverage_count":3},{"rank":6,"config_id":"raw--gemini-2.5-flash--tdefault--tplcurrent--4a7db0f2","provider":"Gemini 2.5 Flash","model":"Gemini 2.5 Flash","provider_id":"raw","model_id":"google/gemini-2.5-flash","dataset":"subpop","framework":"raw","sps":0.782569,"p_dist":0.669404,"p_rank":0.698371,"p_refuse":0.979931,"jsd":0.330596,"tau":0.396743,"n":100,"ci_lower":0.755283,"ci_upper":0.807525,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.771141,"General Attitudes":0.744894,"Health & Science":0.79815,"Identity & Demographics":0.845662,"International Relations & Security":0.633682,"Politics & Governance":0.833952,"Social Values & Religion":0.573939,"Technology & Digital Life":0.698443},"topic_metrics":{"Economy & Work":{"sps":0.771141,"n":9,"p_dist":0.765683,"p_rank":0.776598,"p_refuse":0.990761},"General Attitudes":{"sps":0.744894,"n":18,"p_dist":0.784499,"p_rank":0.70529,"p_refuse":0.937896},"Health & Science":{"sps":0.79815,"n":5,"p_dist":0.783668,"p_rank":0.812633,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.845662,"n":1,"p_dist":0.837771,"p_rank":0.853554,"p_refuse":0.994858},"International Relations & Security":{"sps":0.633682,"n":10,"p_dist":0.594874,"p_rank":0.672489,"p_refuse":0.988524},"Politics & Governance":{"sps":0.833952,"n":4,"p_dist":0.784729,"p_rank":0.883175,"p_refuse":0.98068},"Social Values & Religion":{"sps":0.573939,"n":28,"p_dist":0.543235,"p_rank":0.604642,"p_refuse":0.984966},"Technology & Digital Life":{"sps":0.698443,"n":25,"p_dist":0.674957,"p_rank":0.72193,"p_refuse":0.993706}},"demographic_scorecard":null,"cost_usd":0.0708,"cost_per_100q":0.0708,"cost_per_sps_point":0.090471,"cost_per_response":0.000047,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.675574,"sps_private":0.705267,"sps_public_private_delta":0.029693,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":7,"config_id":"raw--gpt-4o-mini--tdefault--tplcurrent--4dba4f5c","provider":"GPT-4o-mini","model":"GPT-4o-mini","provider_id":"raw","model_id":"openai/gpt-4o-mini","dataset":"subpop","framework":"raw","sps":0.770135,"p_dist":0.628358,"p_rank":0.702118,"p_refuse":0.979931,"jsd":0.371642,"tau":0.404236,"n":100,"ci_lower":0.743473,"ci_upper":0.795293,"is_baseline":false,"is_ensemble":false,"samples_per_question":15,"topic_scores":{"Economy & Work":0.637984,"General Attitudes":0.650164,"Health & Science":0.75095,"Identity & Demographics":0.505814,"International Relations & Security":0.739793,"Politics & Governance":0.751338,"Social Values & Religion":0.671286,"Technology & Digital Life":0.624765},"topic_metrics":{"Economy & Work":{"sps":0.637984,"n":9,"p_dist":0.611267,"p_rank":0.664702,"p_refuse":0.990761},"General Attitudes":{"sps":0.650164,"n":18,"p_dist":0.636684,"p_rank":0.663644,"p_refuse":0.937896},"Health & Science":{"sps":0.75095,"n":5,"p_dist":0.782592,"p_rank":0.719308,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.505814,"n":1,"p_dist":0.393776,"p_rank":0.617851,"p_refuse":0.994858},"International Relations & Security":{"sps":0.739793,"n":10,"p_dist":0.699178,"p_rank":0.780408,"p_refuse":0.988524},"Politics & Governance":{"sps":0.751338,"n":4,"p_dist":0.72769,"p_rank":0.774987,"p_refuse":0.98068},"Social Values & Religion":{"sps":0.671286,"n":28,"p_dist":0.605961,"p_rank":0.736611,"p_refuse":0.984966},"Technology & Digital Life":{"sps":0.624765,"n":25,"p_dist":0.587915,"p_rank":0.661615,"p_refuse":0.993706}},"demographic_scorecard":null,"cost_usd":0.0315,"cost_per_100q":0.0315,"cost_per_sps_point":0.040902,"cost_per_response":0.000021,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.645093,"sps_private":0.717038,"sps_public_private_delta":0.071945,"verification_badge":"flagged","run_count":1,"dataset_coverage_count":3},{"rank":8,"config_id":"raw--claude-haiku-4-5--tdefault--tplcurrent--5fc84438","provider":"Claude Haiku 4.5","model":"Claude Haiku 4.5","provider_id":"raw","model_id":"anthropic/claude-haiku-4-5","dataset":"subpop","framework":"raw","sps":0.76832,"p_dist":0.616186,"p_rank":0.713243,"p_refuse":0.975532,"jsd":0.383814,"tau":0.426486,"n":200,"ci_lower":0.747954,"ci_upper":0.785952,"is_baseline":false,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.654794,"General Attitudes":0.662439,"Health & Science":0.816694,"Identity & Demographics":0.830544,"International Relations & Security":0.652097,"Politics & Governance":0.661457,"Social Values & Religion":0.632518,"Technology & Digital Life":0.685549,"Trust & Wellbeing":0.662224},"topic_metrics":{"Economy & Work":{"sps":0.654794,"n":17,"p_dist":0.602801,"p_rank":0.706786,"p_refuse":0.991415},"General Attitudes":{"sps":0.662439,"n":37,"p_dist":0.625939,"p_rank":0.698939,"p_refuse":0.913336},"Health & Science":{"sps":0.816694,"n":5,"p_dist":0.808985,"p_rank":0.824402,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.830544,"n":1,"p_dist":0.887226,"p_rank":0.773861,"p_refuse":0.994858},"International Relations & Security":{"sps":0.652097,"n":33,"p_dist":0.592516,"p_rank":0.711678,"p_refuse":0.991075},"Politics & Governance":{"sps":0.661457,"n":22,"p_dist":0.594316,"p_rank":0.728598,"p_refuse":0.983152},"Social Values & Religion":{"sps":0.632518,"n":36,"p_dist":0.585516,"p_rank":0.67952,"p_refuse":0.986631},"Technology & Digital Life":{"sps":0.685549,"n":47,"p_dist":0.637244,"p_rank":0.733855,"p_refuse":0.992881},"Trust & Wellbeing":{"sps":0.662224,"n":2,"p_dist":0.620323,"p_rank":0.704124,"p_refuse":0.987774}},"demographic_scorecard":null,"cost_usd":0.864,"cost_per_100q":0.432,"cost_per_sps_point":1.124531,"cost_per_response":0.000144,"tokens_per_response":null,"is_cost_estimated":true,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.651697,"sps_private":0.699911,"sps_public_private_delta":0.048215,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":9,"config_id":"baseline--random-baseline--tdefault--tplcurrent--c53167f6","provider":"Random Baseline","model":"Random Baseline","provider_id":"baseline","model_id":"random-baseline","dataset":"subpop","framework":"baseline","sps":0.757397,"p_dist":0.816139,"p_rank":0.480518,"p_refuse":0.975532,"jsd":0.183861,"tau":-0.038963,"n":200,"ci_lower":0.743615,"ci_upper":0.770987,"is_baseline":true,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.68879,"General Attitudes":0.643775,"Health & Science":0.671718,"Identity & Demographics":0.825967,"International Relations & Security":0.639301,"Politics & Governance":0.623277,"Social Values & Religion":0.639859,"Technology & Digital Life":0.653654,"Trust & Wellbeing":0.693195},"topic_metrics":{"Economy & Work":{"sps":0.68879,"n":17,"p_dist":0.85187,"p_rank":0.525711,"p_refuse":0.991415},"General Attitudes":{"sps":0.643775,"n":37,"p_dist":0.802756,"p_rank":0.484795,"p_refuse":0.913336},"Health & Science":{"sps":0.671718,"n":5,"p_dist":0.806921,"p_rank":0.536515,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.825967,"n":1,"p_dist":0.798381,"p_rank":0.853554,"p_refuse":0.994858},"International Relations & Security":{"sps":0.639301,"n":33,"p_dist":0.798026,"p_rank":0.480575,"p_refuse":0.991075},"Politics & Governance":{"sps":0.623277,"n":22,"p_dist":0.820644,"p_rank":0.425911,"p_refuse":0.983152},"Social Values & Religion":{"sps":0.639859,"n":36,"p_dist":0.828087,"p_rank":0.45163,"p_refuse":0.986631},"Technology & Digital Life":{"sps":0.653654,"n":47,"p_dist":0.813577,"p_rank":0.493731,"p_refuse":0.992881},"Trust & Wellbeing":{"sps":0.693195,"n":2,"p_dist":0.88639,"p_rank":0.5,"p_refuse":0.987774}},"demographic_scorecard":null,"cost_usd":0,"cost_per_100q":0,"cost_per_sps_point":0,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.646199,"sps_private":0.654087,"sps_public_private_delta":0.007889,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3},{"rank":10,"config_id":"baseline--majority-baseline--tdefault--tplcurrent--bd8273b6","provider":"Majority Baseline","model":"Majority Baseline","provider_id":"baseline","model_id":"majority-baseline","dataset":"subpop","framework":"baseline","sps":0.672732,"p_dist":0.466513,"p_rank":0.576152,"p_refuse":0.975532,"jsd":0.533487,"tau":0.152304,"n":200,"ci_lower":0.652763,"ci_upper":0.691308,"is_baseline":true,"is_ensemble":false,"samples_per_question":30,"topic_scores":{"Economy & Work":0.519985,"General Attitudes":0.553876,"Health & Science":0.670485,"Identity & Demographics":0.294864,"International Relations & Security":0.546957,"Politics & Governance":0.530626,"Social Values & Religion":0.5007,"Technology & Digital Life":0.470256,"Trust & Wellbeing":0.717747},"topic_metrics":{"Economy & Work":{"sps":0.519985,"n":17,"p_dist":0.464101,"p_rank":0.575869,"p_refuse":0.991415},"General Attitudes":{"sps":0.553876,"n":37,"p_dist":0.517543,"p_rank":0.590209,"p_refuse":0.913336},"Health & Science":{"sps":0.670485,"n":5,"p_dist":0.570759,"p_rank":0.770212,"p_refuse":0.993911},"Identity & Demographics":{"sps":0.294864,"n":1,"p_dist":0.207578,"p_rank":0.382149,"p_refuse":0.994858},"International Relations & Security":{"sps":0.546957,"n":33,"p_dist":0.488699,"p_rank":0.605214,"p_refuse":0.991075},"Politics & Governance":{"sps":0.530626,"n":22,"p_dist":0.451991,"p_rank":0.609261,"p_refuse":0.983152},"Social Values & Religion":{"sps":0.5007,"n":36,"p_dist":0.448218,"p_rank":0.553181,"p_refuse":0.986631},"Technology & Digital Life":{"sps":0.470256,"n":47,"p_dist":0.415595,"p_rank":0.524916,"p_refuse":0.992881},"Trust & Wellbeing":{"sps":0.717747,"n":2,"p_dist":0.73137,"p_rank":0.704124,"p_refuse":0.987774}},"demographic_scorecard":null,"cost_usd":0,"cost_per_100q":0,"cost_per_sps_point":0,"cost_per_response":null,"tokens_per_response":null,"is_cost_estimated":false,"latency_p50_seconds":null,"latency_p95_seconds":null,"sps_public":0.520528,"sps_private":0.523508,"sps_public_private_delta":0.00298,"verification_badge":"verified","run_count":1,"dataset_coverage_count":3}],"convergence":[{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":5,"sps":0.7467},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":5,"sps":0.7389},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":5,"sps":0.7421},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":5,"sps":0.7495},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.7657},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.7436},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.7414},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.7335},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.7417},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.733},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":10,"sps":0.7358},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":15,"sps":0.5991},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7432},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7452},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7496},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7472},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.6647},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.748},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7368},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7112},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7329},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7279},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7282},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7272},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7242},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7268},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":100,"sps":0.7539},{"model":"openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":100,"sps":0.7533},{"model":"openrouter/google/gemini-2.5-flash","dataset":"opinionsqa","rep_count":10,"sps":0.7511},{"model":"openrouter/google/gemini-2.5-flash","dataset":"opinionsqa","rep_count":15,"sps":0.6658},{"model":"openrouter/google/gemini-2.5-flash","dataset":"opinionsqa","rep_count":15,"sps":0.6839},{"model":"openrouter/google/gemini-2.5-flash","dataset":"opinionsqa","rep_count":50,"sps":0.7306},{"model":"openrouter/google/gemini-2.5-flash","dataset":"opinionsqa","rep_count":50,"sps":0.7491},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":10,"sps":0.756},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":10,"sps":0.742},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":15,"sps":0.6536},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":30,"sps":0.7055},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7512},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7502},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7278},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7337},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7263},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7279},{"model":"openrouter/meta-llama/llama-3.3-70b-instruct","dataset":"opinionsqa","rep_count":50,"sps":0.7275},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":5,"sps":0.7424},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":5,"sps":0.7169},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":10,"sps":0.729},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":10,"sps":0.7663},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":10,"sps":0.744},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":10,"sps":0.7365},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":15,"sps":0.6407},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":15,"sps":0.6652},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.7394},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.7483},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.75},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7451},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7507},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7219},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7197},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.724},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7161},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7161},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":100,"sps":0.7573},{"model":"openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":100,"sps":0.756},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":10,"sps":0.7358},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":30,"sps":0.6555},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":30,"sps":0.7437},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":50,"sps":0.7574},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":50,"sps":0.7771},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":50,"sps":0.764},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":50,"sps":0.7572},{"model":"synthpanel/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":50,"sps":0.764},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.6469},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.6434},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.6446},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7654},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7614},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7612},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7346},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7348},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.7288},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":30,"sps":0.6897},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7656},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7643},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.762},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7842},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7817},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7838},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7841},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7813},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.7889},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5","dataset":"opinionsqa","rep_count":50,"sps":0.6849},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.6915},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.6931},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.6924},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.6924},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.6967},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":30,"sps":0.6988},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7429},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7437},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7419},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7368},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7306},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7362},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7679},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7645},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.763},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7399},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7405},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7407},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7219},{"model":"synthpanel/openrouter/openai/gpt-4o-mini","dataset":"opinionsqa","rep_count":50,"sps":0.7085},{"model":"openrouter/anthropic/claude-sonnet-4","dataset":"opinionsqa","rep_count":10,"sps":0.6384},{"model":"openrouter/anthropic/claude-sonnet-4","dataset":"opinionsqa","rep_count":50,"sps":0.6553},{"model":"openrouter/anthropic/claude-sonnet-4","dataset":"opinionsqa","rep_count":50,"sps":0.6789},{"model":"openrouter/openai/gpt-4o","dataset":"opinionsqa","rep_count":10,"sps":0.7037},{"model":"openrouter/openai/gpt-4o","dataset":"opinionsqa","rep_count":50,"sps":0.7234},{"model":"raw-anthropic/claude-haiku-4-5-20251001","dataset":"opinionsqa","rep_count":10,"sps":0.7147},{"model":"raw-anthropic/claude-haiku-4-5-20251001","dataset":"opinionsqa","rep_count":30,"sps":0.7238},{"model":"raw-gemini/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":10,"sps":0.7619},{"model":"raw-gemini/gemini-2.5-flash-lite","dataset":"opinionsqa","rep_count":30,"sps":0.7378},{"model":"synthpanel/claude-haiku-4-5-20251001","dataset":"opinionsqa","rep_count":3,"sps":0.5497},{"model":"synthpanel/claude-haiku-4-5-20251001","dataset":"opinionsqa","rep_count":10,"sps":0.7781},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.719},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7214},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7159},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7121},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7173},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7221},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7129},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7174},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7546},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":15,"sps":0.7079},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7873},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7828},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7927},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7261},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7296},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7173},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7268},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.723},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7217},{"model":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"opinionsqa","rep_count":30,"sps":0.7246}],"findings":{"sps_convention":"sps","generated_from":"leaderboard-results/ (recomputed per-question rows)","temperature_sweep":[{"model":"Claude Haiku 4.5","temperature":0.3,"sps":0.842561,"n_runs":3,"dataset":"opinionsqa","std":0.002891},{"model":"Claude Haiku 4.5","temperature":0.5,"sps":0.843958,"n_runs":3,"dataset":"opinionsqa","std":0.003139},{"model":"Claude Haiku 4.5","temperature":0.7,"sps":0.844639,"n_runs":3,"dataset":"opinionsqa","std":0.006739},{"model":"Claude Haiku 4.5","temperature":0.85,"sps":0.850292,"n_runs":3,"dataset":"opinionsqa","std":0.006396},{"model":"Claude Haiku 4.5","temperature":1,"sps":0.844267,"n_runs":3,"dataset":"opinionsqa","std":0.007317},{"model":"GPT-4o-mini","temperature":0.3,"sps":0.816834,"n_runs":2,"dataset":"opinionsqa","std":0.000275},{"model":"GPT-4o-mini","temperature":0.5,"sps":0.821763,"n_runs":2,"dataset":"opinionsqa","std":0.000741},{"model":"GPT-4o-mini","temperature":0.7,"sps":0.822453,"n_runs":2,"dataset":"opinionsqa","std":0.000225},{"model":"GPT-4o-mini","temperature":0.85,"sps":0.823682,"n_runs":2,"dataset":"opinionsqa","std":0.001619},{"model":"GPT-4o-mini","temperature":1,"sps":0.828525,"n_runs":2,"dataset":"opinionsqa","std":0.001244},{"model":"Gemini Flash Lite","temperature":0.3,"sps":0.8189,"n_runs":2,"dataset":"opinionsqa","std":0.002972},{"model":"Gemini Flash Lite","temperature":0.5,"sps":0.826995,"n_runs":2,"dataset":"opinionsqa","std":0.001454},{"model":"Gemini Flash Lite","temperature":0.7,"sps":0.837993,"n_runs":2,"dataset":"opinionsqa","std":0.005218},{"model":"Gemini Flash Lite","temperature":0.85,"sps":0.8486,"n_runs":2,"dataset":"opinionsqa","std":0.002597},{"model":"Gemini Flash Lite","temperature":1,"sps":0.856044,"n_runs":2,"dataset":"opinionsqa","std":0.002852},{"model":"Gemini Flash Lite","temperature":1.2,"sps":0.85561,"n_runs":2,"dataset":"opinionsqa","std":0.001283},{"model":"Gemini Flash Lite","temperature":1.5,"sps":0.858208,"n_runs":2,"dataset":"opinionsqa","std":0.000433},{"model":"Gemini Flash Lite","temperature":1.8,"sps":0.856798,"n_runs":2,"dataset":"opinionsqa","std":0.005319},{"model":"Gemini Flash Lite","temperature":2,"sps":0.86376,"n_runs":2,"dataset":"opinionsqa","std":0.000782}],"ensemble_comparison":[{"dataset":"globalopinionqa","best_single_model":"SynthPanel (GPT-4o-mini)","best_single_framework":"product","best_single_sps":0.786229,"ensemble_sps":0.81346,"improvement":0.027231,"n_questions":100,"random_baseline_sps":0.709683,"random_baseline_n":10},{"dataset":"opinionsqa","best_single_model":"Gemini 2.5 Flash","best_single_framework":"raw","best_single_sps":0.829212,"ensemble_sps":0.876721,"improvement":0.047509,"n_questions":684,"random_baseline_sps":0.762852,"random_baseline_n":684},{"dataset":"subpop","best_single_model":"SynthPanel (Gemini Flash Lite)","best_single_framework":"product","best_single_sps":0.820993,"ensemble_sps":0.830664,"improvement":0.009671,"n_questions":200,"random_baseline_sps":0.757397,"random_baseline_n":200}],"conditioning_results":[{"attribute":"POLPARTY","group":"Republican","group_raw":"Republican","p_dist":0.666078,"p_cond":0.072785,"n_replications":4,"p_cond_std":0.003826},{"attribute":"POLPARTY","group":"Democrat","group_raw":"Democrat","p_dist":0.643987,"p_cond":0.032835,"n_replications":4,"p_cond_std":0.005362},{"attribute":"INCOME","group":"$100K+","group_raw":"$100,000 or more","p_dist":0.673734,"p_cond":0.029827,"n_replications":3,"p_cond_std":0.003656},{"attribute":"INCOME","group":"<$30K","group_raw":"Less than $30,000","p_dist":0.595654,"p_cond":0.019292,"n_replications":3,"p_cond_std":0.003217},{"attribute":"EDUCATION","group":"Less than HS","group_raw":"Less than high school","p_dist":0.596863,"p_cond":0.037923,"n_replications":1},{"attribute":"EDUCATION","group":"College graduate","group_raw":"College graduate/some postgrad","p_dist":0.640846,"p_cond":0.035699,"n_replications":1}],"conditioning_extended":[{"attribute":"CREGION","group":"South","group_raw":"South","p_dist":0.640807,"p_cond":0.039304,"n_replications":2,"p_cond_std":0.004909},{"attribute":"CREGION","group":"Northeast","group_raw":"Northeast","p_dist":0.628108,"p_cond":0.021031,"n_replications":2,"p_cond_std":0.003718},{"attribute":"RACE","group":"White","group_raw":"White","p_dist":0.78508,"p_cond":0.140759,"n_replications":3,"p_cond_std":0.0303},{"attribute":"RACE","group":"Hispanic","group_raw":"Hispanic","p_dist":0.605304,"p_cond":0.033923,"n_replications":3,"p_cond_std":0.004766},{"attribute":"RACE","group":"Asian","group_raw":"Asian","p_dist":0.636558,"p_cond":0.025403,"n_replications":3,"p_cond_std":0.005113},{"attribute":"RACE","group":"Black","group_raw":"Black","p_dist":0.584576,"p_cond":0.02457,"n_replications":3,"p_cond_std":0.00274},{"attribute":"RELIG","group":"Protestant","group_raw":"Protestant","p_dist":0.838025,"p_cond":0.178139,"n_replications":1},{"attribute":"RELIG","group":"Muslim","group_raw":"Muslim","p_dist":0.819781,"p_cond":0.161155,"n_replications":1},{"attribute":"RELIG","group":"Hindu","group_raw":"Hindu","p_dist":0.806252,"p_cond":0.129162,"n_replications":1},{"attribute":"RELIG","group":"Jewish","group_raw":"Jewish","p_dist":0.773676,"p_cond":0.126971,"n_replications":1},{"attribute":"RELIG","group":"Atheist","group_raw":"Atheist","p_dist":0.73835,"p_cond":0.09528,"n_replications":1},{"attribute":"SEX","group":"Male","group_raw":"Male","p_dist":0.647331,"p_cond":0.054629,"n_replications":1},{"attribute":"SEX","group":"Female","group_raw":"Female","p_dist":0.617469,"p_cond":0.026534,"n_replications":1}],"template_comparison":[{"template":"current","sps":0.690325,"n_runs":2,"std":0.019021},{"template":"minimal","sps":0.580775,"n_runs":2,"std":0.005248},{"template":"demo","sps":0.569368,"n_runs":2,"std":0.000917},{"template":"values","sps":0.554686,"n_runs":2,"std":0.032431}],"lever_hierarchy":[{"name":"Ensemble blending","effect_min":1,"effect_max":4.8,"cost":"zero","status":"done"},{"name":"Per-model optimal temperature","effect_min":0.8,"effect_max":4.5,"cost":"low","status":"actionable"},{"name":"Demographic conditioning","effect_min":1.9,"effect_max":7.3,"cost":"moderate","status":"scientific"},{"name":"Persona template","effect_min":0,"effect_max":0,"cost":"zero","status":"done","note":"'current' template already optimal (+11.0 pts over the best alternative); no further gain available from this lever."}],"comparison_sets":{"ensemble_comparison":"best_single = highest recomputed-SPS non-ensemble, non-baseline leaderboard row (raw or product framework) evaluated on at least as many questions as the ensemble, after per-(model, framework, dataset) dedup; random_baseline = recomputed SPS of the random-baseline run for the same dataset (n in random_baseline_n).","temperature_sweep":"OpinionsQA sweep-tagged SynthPanel runs (' t=' provider suffix); mean/std of recomputed SPS across the n_runs replications per (model, temperature) cell.","conditioning_results":"Per-group conditioned SubPOP evaluations (Haiku 4.5, t=0.85); mean p_dist / p_cond across n_replications runs carrying a demographic_breakdown for the group.","template_comparison":"SubPOP persona-template variant runs (Haiku 4.5, t=0.85, 100 questions); mean/std of recomputed SPS across n_runs."},"asserted_constants":[{"name":"blend_weighting_equivalence","value":"equal-weight ~= score-proportional ~= inverse-JSD (to 3 decimals)","source":"Session-1 ensemble weighting comparison (2026-04-12); the alternative-weighting blend files were not committed to leaderboard-results/, so this is not reproducible from artifacts."},{"name":"per_question_improvement_share","value":"72-81% of individual questions improve under blending","source":"Session-1 per-question blend analysis (2026-04-12), computed under the retired parity-2 convention; not recomputed since."},{"name":"oracle_ceiling_gap","value":"per-question oracle model selection barely exceeds the equal blend","source":"Session-1 oracle analysis (2026-04-12); oracle blend files were not committed to leaderboard-results/."},{"name":"base_entropy_h5","value":"base output entropy (bits, default temperature): GPT-4o-mini 0.22, Claude Haiku 4.5 0.36, Gemini Flash Lite 0.56","source":"Experiment H5 notebook analysis (2026-04-12) of default-temperature model_distribution rows; the notebook output was not committed, so the exact values are asserted."},{"name":"template_refusal_collapse","value":"P_refuse collapses from ~0.80 to 0.40-0.50 on templates with unfilled format-string placeholders","source":"Session-1 template-variant analysis (2026-04-12); per-component P_refuse for the template runs is derivable in principle but the published block only tracks composite SPS for templates."}],"caveats":[]},"baselines":{"ceiling":{},"temporal_drift":null,"citations":[{"key":"santurkar2023","text":"Santurkar et al. 2023, OpinionsQA.","arxiv":"2303.17548"},{"key":"durmus2023","text":"Durmus et al. 2023, GlobalOpinionQA.","arxiv":"2306.16388"},{"key":"suh2025","text":"Suh et al. 2025, SubPOP.","arxiv":"2502.16761"},{"key":"pew_methods","text":"Pew Research Center, Survey Methodology 101."},{"key":"spearman1910","text":"Spearman, C. (1910); Brown, W. (1910). Spearman-Brown prophecy formula."},{"key":"efron1979","text":"Efron, B. (1979). Bootstrap methods."},{"key":"lin1991","text":"Lin, J. (1991). Jensen-Shannon divergence properties."},{"key":"cochran1977","text":"Cochran, W. (1977). Sampling Techniques (n=400 rule-of-thumb)."}],"survey_weight_caveat":"Ceiling computed from raw category counts; ignoring survey weights could shift the ceiling by 1-3% on demographically skewed subgroups."},"cross_provider_concordance":{"globalopinionqa":{"models":["Claude Haiku 4.5","Claude Sonnet 4.6","GPT-4o-mini","Gemini 2.5 Flash","Llama 3.3 70B"],"matrix":[[0,0.335683,0.27869,0.269583,0.311335],[0.335683,0,0.324039,0.326184,0.275544],[0.27869,0.324039,0,0.277088,0.21221],[0.269583,0.326184,0.277088,0,0.282885],[0.311335,0.275544,0.21221,0.282885,0]],"mean_cross_model_jsd":0.289324,"mean_human_jsd":0.370146},"opinionsqa":{"models":["Claude Haiku 4.5","Claude Sonnet 4","GPT-4o","GPT-4o-mini","Gemini 2.5 Flash","Gemini Flash Lite","Llama 3.3 70B"],"matrix":[[0,0.254674,0.29902,0.303531,0.297094,0.309816,0.273953],[0.254674,0,0.336434,0.402281,0.354759,0.403376,0.379415],[0.29902,0.336434,0,0.161905,0.219506,0.228906,0.229255],[0.303531,0.402281,0.161905,0,0.219523,0.195565,0.204973],[0.297094,0.354759,0.219506,0.219523,0,0.218144,0.254065],[0.309816,0.403376,0.228906,0.195565,0.218144,0,0.342674],[0.273953,0.379415,0.229255,0.204973,0.254065,0.342674,0]],"mean_cross_model_jsd":0.280422,"mean_human_jsd":0.297788},"subpop":{"models":["Claude Haiku 4.5","GPT-4o-mini","Gemini 2.5 Flash","Llama 3.3 70B"],"matrix":[[0,0.381428,0.24479,0.303028],[0.381428,0,0.330614,0.31692],[0.24479,0.330614,0,0.30062],[0.303028,0.31692,0.30062,0]],"mean_cross_model_jsd":0.3129,"mean_human_jsd":0.357738}},"pricing_snapshot":{"generated_at":"2026-07-14T23:20:52.886103+00:00","synth_panel_version":"1.5.7","snapshot_date":"2026-05-10","rates":{"haiku":{"input_cost_per_million":1,"output_cost_per_million":5,"cache_creation_cost_per_million":1.25,"cache_read_cost_per_million":0.1},"sonnet":{"input_cost_per_million":3,"output_cost_per_million":15,"cache_creation_cost_per_million":3.75,"cache_read_cost_per_million":0.3},"opus":{"input_cost_per_million":15,"output_cost_per_million":75,"cache_creation_cost_per_million":18.75,"cache_read_cost_per_million":1.5},"gemini-2.5-pro":{"input_cost_per_million":1.25,"output_cost_per_million":10,"cache_creation_cost_per_million":0.375,"cache_read_cost_per_million":0.125},"gemini-flash":{"input_cost_per_million":0.3,"output_cost_per_million":2.5,"cache_creation_cost_per_million":0.0833,"cache_read_cost_per_million":0.03}}},"dataset_policies":[{"name":"eurobarometer","redistribution_policy":"gated","license_url":"https://www.gesis.org/eurobarometer-data-service/about-eurobarometer","citation":"European Commission / GESIS — Eurobarometer Consumer Modules"},{"name":"globalopinionqa","redistribution_policy":"gated","license_url":"https://creativecommons.org/licenses/by-nc-sa/4.0/","citation":"Durmus et al. 2023, Anthropic — llm_global_opinions"},{"name":"gss","redistribution_policy":"full","license_url":"https://gss.norc.org/About-The-GSS","citation":"NORC at the University of Chicago — General Social Survey"},{"name":"michigan","redistribution_policy":"gated","license_url":"https://data.sca.isr.umich.edu/terms.php","citation":"Survey Research Center, University of Michigan — Surveys of Consumers"},{"name":"ntia","redistribution_policy":"full","license_url":"https://www.ntia.gov/page/download-ntia-internet-use-survey-datasets","citation":"NTIA Internet Use Survey (U.S. Government work, 17 USC 105)"},{"name":"opinionsqa","redistribution_policy":"gated","license_url":"https://worksheets.codalab.org/worksheets/0x6fb693719477478aac73fc07db333f69","citation":"Santurkar et al., ICML 2023 — Whose Opinions Do LLMs Reflect? (derived from Pew American Trends Panel)"},{"name":"pewtech","redistribution_policy":"gated","license_url":"https://www.pewresearch.org/internet/datasets/","citation":"Pew Research Center — Internet & Technology (American Trends Panel)"},{"name":"subpop","redistribution_policy":"gated","license_url":"https://huggingface.co/datasets/jjssuh/subpop","citation":"Suh et al., ACL 2025 — SubPOP: Subpopulation-Level Opinion Prediction"},{"name":"wvs","redistribution_policy":"gated","license_url":"https://worldvaluessurvey.org/WVSDocumentationWV7.jsp","citation":"World Values Survey Association — WVS Wave 7 (2017-2022)"}],"excluded_runs":[{"run_id":"globalopinionqa_synthpanel_openrouter_anthropic_claude-sonnet-4_20260411_230741","reason":"uniform-garbage: 100/100 questions (100.0%) have uniform model_distribution with mean refusal_rate 0.000","provider":"synthpanel/openrouter/anthropic/claude-sonnet-4","dataset":"globalopinionqa","samples_per_question":30,"n_evaluated":100,"timestamp":"2026-04-12T04:07:41.786580+00:00","metrics":{"n_questions":100,"n_uniform_questions":100,"uniform_fraction":1,"refusal_rate":0,"n_zero_sample_questions":0,"zero_sample_fraction":0}},{"run_id":"globalopinionqa_synthpanel_openrouter_anthropic_claude-sonnet-4_20260411_230957","reason":"uniform-garbage: 100/100 questions (100.0%) have uniform model_distribution with mean refusal_rate 0.000","provider":"synthpanel/openrouter/anthropic/claude-sonnet-4","dataset":"globalopinionqa","samples_per_question":30,"n_evaluated":100,"timestamp":"2026-04-12T04:09:57.466611+00:00","metrics":{"n_questions":100,"n_uniform_questions":100,"uniform_fraction":1,"refusal_rate":0,"n_zero_sample_questions":0,"zero_sample_fraction":0}},{"run_id":"subpop_synthpanel_claude-haiku-4-5-20251001_20260411_073013","reason":"uniform-garbage: 160/200 questions (80.0%) have uniform model_distribution with mean refusal_rate 0.012","provider":"synthpanel/claude-haiku-4-5-20251001","dataset":"subpop","samples_per_question":30,"n_evaluated":200,"timestamp":"2026-04-11T12:30:13.696991+00:00","metrics":{"n_questions":200,"n_uniform_questions":160,"uniform_fraction":0.8,"refusal_rate":0.011999995000000001,"n_zero_sample_questions":0,"zero_sample_fraction":0}},{"run_id":"subpop_synthpanel_openrouter_anthropic_claude-haiku-4-5 t=0.85_20260412_030546","reason":"uniform-garbage: 100/100 questions (100.0%) have uniform model_distribution with mean refusal_rate 0.000","provider":"synthpanel/openrouter/anthropic/claude-haiku-4-5 t=0.85","dataset":"subpop","samples_per_question":15,"n_evaluated":100,"timestamp":"2026-04-12T08:05:46.652957+00:00","metrics":{"n_questions":100,"n_uniform_questions":100,"uniform_fraction":1,"refusal_rate":0,"n_zero_sample_questions":0,"zero_sample_fraction":0}},{"run_id":"subpop_synthpanel_openrouter_anthropic_claude-sonnet-4_20260411_230420","reason":"uniform-garbage: 200/200 questions (100.0%) have uniform model_distribution with mean refusal_rate 0.000","provider":"synthpanel/openrouter/anthropic/claude-sonnet-4","dataset":"subpop","samples_per_question":30,"n_evaluated":200,"timestamp":"2026-04-12T04:04:21.147881+00:00","metrics":{"n_questions":200,"n_uniform_questions":200,"uniform_fraction":1,"refusal_rate":0,"n_zero_sample_questions":0,"zero_sample_fraction":0}},{"run_id":"subpop_synthpanel_openrouter_anthropic_claude-sonnet-4_20260411_230525","reason":"uniform-garbage: 200/200 questions (100.0%) have uniform model_distribution with mean refusal_rate 0.000","provider":"synthpanel/openrouter/anthropic/claude-sonnet-4","dataset":"subpop","samples_per_question":30,"n_evaluated":200,"timestamp":"2026-04-12T04:05:26.684148+00:00","metrics":{"n_questions":200,"n_uniform_questions":200,"uniform_fraction":1,"refusal_rate":0,"n_zero_sample_questions":0,"zero_sample_fraction":0}}]}