General QA Verification on TriviaQA
0.9086Precision@1DiVA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DiVABackbone=Qwen-2.5-7B-Instruct, AR=true, EK=true, FS=true, TE=true2026.01 | 0.9086 | 0.8853 | |
| DiVABackbone=Qwen-3-4B-Instruct, AR=true, EK=true, FS=true, TE=true2026.01 | 0.8898 | 0.8423 | |
| DiVA w/o compressionBackbone=Qwen-3-4B-Instruct, AR=true, EK=true, FS=true, TE=true2026.01 | 0.879 | 0.8369 | |
| DiVA w/o compressionBackbone=Qwen-2.5-7B-Instruct, AR=true, EK=true, FS=true, TE=true2026.01 | 0.8387 | 0.7939 | |
| Agentic-Generative VerifierBackbone=Qwen-3-4B-Instruct, AR=true, EK=true, FS=false, TE=true2026.01 | 0.8199 | 0.7616 | |
| Agentic-Generative VerifierBackbone=Qwen-2.5-7B-Instruct, AR=true, EK=true, FS=false, TE=true2026.01 | 0.8091 | 0.7473 | |
| Agentic-Generative ScorerBackbone=Qwen-2.5-7B-Instruct, AR=true, EK=true, FS=true, TE=false2026.01 | 0.7898 | 0.7269 | |
| Agentic-Generative ScorerBackbone=Qwen-3-4B-Instruct, AR=true, EK=true, FS=true, TE=false2026.01 | 0.7898 | 0.7269 | |
| Discriminative VerifierBackbone=Qwen-2.5-7B-Instruct, AR=false, EK=false, FS=true, TE=true2026.01 | 0.7581 | 0.638 | |
| MiniCheck w/ MiniCheck-7BBackbone=MiniCheck-7B2026.01 | 0.7446 | 0.6989 | |
| Discriminative VerifierBackbone=Qwen-3-4B-Instruct, AR=false, EK=false, FS=true, TE=true2026.01 | 0.7204 | 0.5484 | |
| Generative VerifierBackbone=Qwen-3-4B-Instruct, AR=true, EK=false, FS=false, TE=true2026.01 | 0.7019 | 0.5896 | |
| FactScore w/ GPT-4Backbone=GPT-42026.01 | 0.7 | 0.7011 | |
| Generative VerifierBackbone=Qwen-2.5-7B-Instruct, AR=true, EK=false, FS=false, TE=true2026.01 | 0.6981 | 0.6137 | |
| HHEM-2.1-Open2026.01 | 0.6935 | 0.6004 | |
| MiniCheck w/ Flan-T5Backbone=Flan-T52026.01 | 0.6882 | 0.6219 |