Fact Verification on POLITIFACT HIDDEN (test)
83.6AccuracyRADARmulti
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| RADARmultiEvidence Setting=Full Evidence, Backbone=GPT-4o-mini, Agent Interaction=multi-agent2026.04 | 83.6 | 70.8 | 59.3 | 61.1 | 92 | |
| CoT + RAEvidence Setting=Full Evidence, Backbone=GPT-4o-mini, TRACER-enhanced=true2026.04 | 78.5 | 68 | 56.7 | 60.2 | 87.1 | |
| HiSS + RAEvidence Setting=Full Evidence, Backbone=GPT-4o-mini, TRACER-enhanced=true2026.04 | 77.8 | 65.7 | 49.1 | 61.5 | 86.5 | |
| RADARmultiEvidence Setting=Retrieved Evidence, Backbone=GPT-4o-mini, Agent Interaction=multi-agent2026.04 | 77.7 | 63.3 | 45.6 | 56.5 | 87.6 | |
| CoTEvidence Setting=Full Evidence, Backbone=GPT-4o-mini2026.04 | 76.3 | 64.3 | 52.9 | 52.8 | 87.1 | |
| HiSSEvidence Setting=Full Evidence, Backbone=GPT-4o-mini2026.04 | 74.7 | 61.9 | 48.1 | 51.2 | 86.4 | |
| RADARsingleEvidence Setting=Full Evidence, Backbone=GPT-4o-mini, Agent Interaction=single-agent2026.04 | 68.7 | 62.4 | 57.6 | 51.1 | 78.6 | |
| D2DEvidence Setting=Retrieved Evidence, Backbone=GPT-4o-mini2026.04 | 63 | 50.9 | 36.3 | 39.7 | 76.5 | |
| FIREEvidence Setting=Retrieved Evidence, Backbone=GPT-4o-mini2026.04 | 60.3 | 46.9 | 30.7 | 34.1 | 75.9 | |
| RADARsingleEvidence Setting=Retrieved Evidence, Backbone=GPT-4o-mini, Agent Interaction=single-agent2026.04 | 58.4 | 51 | 41.2 | 41.5 | 70.2 |