Generation Correctness Prediction on TriviaQA (test)
42.41AURCINSIDE
Evaluation Results
| Method | Links | |
|---|---|---|
| INSIDEBackbone=Qwen3-Instruct4B2026.05 | 42.41 | |
| S.E.Backbone=Qwen3-Instruct4B2026.05 | 40.34 | |
| M.I.Backbone=Qwen3-Instruct4B2026.05 | 37.34 | |
| Self-ConBackbone=Qwen3-Instruct4B2026.05 | 37.32 | |
| P(True)Backbone=Qwen3-Instruct4B2026.05 | 36.51 | |
| S.D.Backbone=Qwen3-Instruct4B2026.05 | 36.48 | |
| DegBackbone=Qwen3-Instruct4B2026.05 | 35.91 | |
| LN-PEBackbone=Qwen3-Instruct4B2026.05 | 33.69 | |
| SARBackbone=Qwen3-Instruct4B2026.05 | 30.84 | |
| SemGradBackbone=Qwen3-Instruct4B2026.05 | 30.77 | |
| ExGradBackbone=Qwen3-Instruct4B2026.05 | 30.73 | |
| G-NLLBackbone=Qwen3-Instruct4B2026.05 | 30.54 | |
| HybridGradBackbone=Qwen3-Instruct4B2026.05 | 30.27 | |
| ParaGradBackbone=Qwen3-Instruct4B2026.05 | 30.14 | |
| INSIDEBackbone=Mistral-Nemo-Instruct12B2026.05 | 18.51 | |
| S.E.Backbone=Mistral-Nemo-Instruct12B2026.05 | 17.77 | |
| INSIDEBackbone=Llama3.1-Instruct8B2026.05 | 16.48 | |
| S.D.Backbone=Mistral-Nemo-Instruct12B2026.05 | 16.34 | |
| Self-ConBackbone=Mistral-Nemo-Instruct12B2026.05 | 15.82 | |
| S.E.Backbone=Llama3.1-Instruct8B2026.05 | 15.47 | |
| P(True)Backbone=Llama3.1-Instruct8B2026.05 | 15.1 | |
| Self-ConBackbone=Llama3.1-Instruct8B2026.05 | 14.92 | |
| DegBackbone=Mistral-Nemo-Instruct12B2026.05 | 14.7 | |
| M.I.Backbone=Mistral-Nemo-Instruct12B2026.05 | 14.69 | |
| P(True)Backbone=Mistral-Nemo-Instruct12B2026.05 | 14.21 | |
| M.I.Backbone=Llama3.1-Instruct8B2026.05 | 14.19 | |
| S.D.Backbone=Llama3.1-Instruct8B2026.05 | 14.02 | |
| DegBackbone=Llama3.1-Instruct8B2026.05 | 13.44 | |
| LN-PEBackbone=Mistral-Nemo-Instruct12B2026.05 | 13.27 | |
| SemGradBackbone=Mistral-Nemo-Instruct12B2026.05 | 12.96 | |
| LN-PEBackbone=Llama3.1-Instruct8B2026.05 | 12.64 | |
| HybridGradBackbone=Mistral-Nemo-Instruct12B2026.05 | 12.27 | |
| ExGradBackbone=Mistral-Nemo-Instruct12B2026.05 | 12.2 | |
| G-NLLBackbone=Mistral-Nemo-Instruct12B2026.05 | 12.14 | |
| SARBackbone=Mistral-Nemo-Instruct12B2026.05 | 12 | |
| ExGradBackbone=Llama3.1-Instruct8B2026.05 | 11.93 | |
| SemGradBackbone=Llama3.1-Instruct8B2026.05 | 11.91 | |
| SARBackbone=Llama3.1-Instruct8B2026.05 | 11.77 | |
| ParaGradBackbone=Mistral-Nemo-Instruct12B2026.05 | 11.72 | |
| G-NLLBackbone=Llama3.1-Instruct8B2026.05 | 11.71 | |
| ParaGradBackbone=Llama3.1-Instruct8B2026.05 | 11.51 | |
| HybridGradBackbone=Llama3.1-Instruct8B2026.05 | 11.5 |