Fact-checking on LIAR
79AccuracyIO
Evaluation Results
| Method | Links | |
|---|---|---|
| IOExecutor=GPT-52026.01 | 79 | |
| CoTExecutor=GPT-52026.01 | 79 | |
| UPAExecutor=GPT-52026.01 | 78.8 | |
| SPOExecutor=GPT-52026.01 | 78 | |
| UPAExecutor=Claude-4.5-Sonnet2026.01 | 74.7 | |
| SPOExecutor=Claude-4.5-Sonnet2026.01 | 74.1 | |
| CoTExecutor=Claude-4.5-Sonnet2026.01 | 73.8 | |
| IOExecutor=Claude-4.5-Sonnet2026.01 | 73.7 | |
| UPAExecutor=DeepSeek-V3.22026.01 | 68.6 | |
| CoTExecutor=DeepSeek-V3.22026.01 | 68.4 | |
| SPOExecutor=DeepSeek-V3.22026.01 | 68.2 | |
| ToEBackbone=GPT-OSS-20B2026.06 | 68 | |
| IOExecutor=DeepSeek-V3.22026.01 | 67.9 | |
| ToEBackbone=DeepSeek-V3.22026.06 | 62 | |
| TELLERBackbone=GPT-OSS-20B2026.06 | 60 | |
| DefGenBackbone=DeepSeek-V3.22026.06 | 50 | |
| STEELBackbone=DeepSeek-V3.22026.06 | 46 | |
| TELLERBackbone=DeepSeek-V3.22026.06 | 46 | |
| AdsentBackbone=GPT-OSS-20B2026.06 | 44 | |
| ZCoTBackbone=DeepSeek-V3.22026.06 | 42 | |
| AdsentBackbone=DeepSeek-V3.22026.06 | 40 | |
| DirectBackbone=DeepSeek-V3.22026.06 | 38 | |
| DefGenBackbone=GPT-OSS-20B2026.06 | 38 | |
| ZCoTBackbone=GPT-OSS-20B2026.06 | 34 | |
| AFacTABackbone=DeepSeek-V3.22026.06 | 32 | |
| AFacTABackbone=GPT-OSS-20B2026.06 | 32 | |
| DirectBackbone=GPT-OSS-20B2026.06 | 24 | |
| STEELBackbone=GPT-OSS-20B2026.06 | 20 |