# Fact Verification on FEVER (EM, $Δ$NF) Source: [Fact Verification on FEVER (EM, $Δ$NF)]() Collection index: [llms.txt]() Data updated: 2026-06-25T08:00:21.023Z Evaluation of model performance on the FEVER dataset for Fact Verification, reporting Exact Match (EM) and Delta Non-Fact ($Δ$NF) scores. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Fact Verification | | Dataset (including any stated split) | FEVER | | Metric | Exact Match (EM) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 30 of 30 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Exact Match (EM)](): Higher is better; unit: PERCENT. - [Delta Non-Fact ($Δ$NF)](): Lower is better; unit: SCALAR. ## Recorded results | Method | Exact Match (EM) | Result date | Report paper | | --- | --- | --- | --- | | MADARA | 93.6 | Not recorded | 6a3cb63c206542251d7ca6a8 | | CoT | 92.6 | Not recorded | 6a3cb63c206542251d7ca6a8 | | MADARA | 92.6 | Not recorded | 6a3cb63c206542251d7ca6a8 | | CoT | 92.5 | Not recorded | 6a3cb63c206542251d7ca6a8 | | SDA | 92.5 | Not recorded | 6a3cb63c206542251d7ca6a8 | | 3-Agent | 92.4 | Not recorded | 6a3cb63c206542251d7ca6a8 | | NF | 92.2 | Not recorded | 6a3cb63c206542251d7ca6a8 | | MADARA (Average) | 91.2 | Not recorded | 6a3cb63c206542251d7ca6a8 | | MADARA | 90.9 | Not recorded | 6a3cb63c206542251d7ca6a8 | | 3-Agent | 90.9 | Not recorded | 6a3cb63c206542251d7ca6a8 | | MADARA | 90.9 | Not recorded | 6a3cb63c206542251d7ca6a8 | | NF | 90.7 | Not recorded | 6a3cb63c206542251d7ca6a8 | | SDA | 90.7 | Not recorded | 6a3cb63c206542251d7ca6a8 | | SDA | 90.5 | Not recorded | 6a3cb63c206542251d7ca6a8 | | Average | 90.4 | Not recorded | 6a3cb63c206542251d7ca6a8 | | Average | 90.1 | Not recorded | 6a3cb63c206542251d7ca6a8 | | SDA | 90 | Not recorded | 6a3cb63c206542251d7ca6a8 | | Average | 90 | Not recorded | 6a3cb63c206542251d7ca6a8 | | 3-Agent | 89.8 | Not recorded | 6a3cb63c206542251d7ca6a8 | | Average | 89.7 | Not recorded | 6a3cb63c206542251d7ca6a8 | | NF | 89.6 | Not recorded | 6a3cb63c206542251d7ca6a8 | | CoT | 89.5 | Not recorded | 6a3cb63c206542251d7ca6a8 | | 3-Agent | 88.9 | Not recorded | 6a3cb63c206542251d7ca6a8 | | NF | 88.7 | Not recorded | 6a3cb63c206542251d7ca6a8 | | CoT | 88.4 | Not recorded | 6a3cb63c206542251d7ca6a8 | | SDA | 88.2 | Not recorded | 6a3cb63c206542251d7ca6a8 | | MADARA | 88.2 | Not recorded | 6a3cb63c206542251d7ca6a8 | | 3-Agent | 87.8 | Not recorded | 6a3cb63c206542251d7ca6a8 | | CoT | 87.5 | Not recorded | 6a3cb63c206542251d7ca6a8 | | NF | 87.1 | Not recorded | 6a3cb63c206542251d7ca6a8 | ### MADARA Exact Match (EM): 93.6. Result ID: 6a3ce002fbc373592c51e4c3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemma-2-9B","RSC":"Quality-Ordered","ρ̂1":"0.58","Strategy":"ATF"}. This record does not assert independent reproduction. ### CoT Exact Match (EM): 92.6. Result ID: 6a3ce002fbc373592c51e4b2. [Reporting paper]() Recorded evaluation configuration: {"Model":"Mistral-7B-v0.3"}. This record does not assert independent reproduction. ### MADARA Exact Match (EM): 92.6. Result ID: 6a3ce002fbc373592c51e4b4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Mistral-7B-v0.3","RSC":"Quality-Ordered","ρ̂1":"0.28","Strategy":"CoT"}. This record does not assert independent reproduction. ### CoT Exact Match (EM): 92.5. Result ID: 6a3ce002fbc373592c51e4c1. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemma-2-9B"}. This record does not assert independent reproduction. ### SDA Exact Match (EM): 92.5. Result ID: 6a3ce002fbc373592c51e4c2. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemma-2-9B"}. This record does not assert independent reproduction. ### 3-Agent Exact Match (EM): 92.4. Result ID: 6a3ce002fbc373592c51e4c0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemma-2-9B"}. This record does not assert independent reproduction. ### NF Exact Match (EM): 92.2. Result ID: 6a3ce002fbc373592c51e4bf. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemma-2-9B"}. This record does not assert independent reproduction. ### MADARA (Average) Exact Match (EM): 91.2. Result ID: 6a3ce002fbc373592c51e4c8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MADARA Exact Match (EM): 90.9. Result ID: 6a3ce002fbc373592c51e4af. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","RSC":"Quality-Ordered","ρ̂1":"0.65","Strategy":"ATF"}. This record does not assert independent reproduction. ### 3-Agent Exact Match (EM): 90.9. Result ID: 6a3ce002fbc373592c51e4b1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Mistral-7B-v0.3"}. This record does not assert independent reproduction. ### MADARA Exact Match (EM): 90.9. Result ID: 6a3ce002fbc373592c51e4b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B","RSC":"Quality-Ordered","ρ̂1":"0.63","Strategy":"ATF"}. This record does not assert independent reproduction. ### NF Exact Match (EM): 90.7. Result ID: 6a3ce002fbc373592c51e4b0. [Reporting paper]() Recorded evaluation configuration: {"Model":"Mistral-7B-v0.3"}. This record does not assert independent reproduction. ### SDA Exact Match (EM): 90.7. Result ID: 6a3ce002fbc373592c51e4b3. [Reporting paper]() Recorded evaluation configuration: {"Model":"Mistral-7B-v0.3"}. This record does not assert independent reproduction. ### SDA Exact Match (EM): 90.5. Result ID: 6a3ce002fbc373592c51e4bd. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### Average Exact Match (EM): 90.4. Result ID: 6a3ce002fbc373592c51e4c7. [Reporting paper]() Recorded evaluation configuration: {"Method\_Group":"SDA"}. This record does not assert independent reproduction. ### Average Exact Match (EM): 90.1. Result ID: 6a3ce002fbc373592c51e4c6. [Reporting paper]() Recorded evaluation configuration: {"Method\_Group":"CoT"}. This record does not assert independent reproduction. ### SDA Exact Match (EM): 90. Result ID: 6a3ce002fbc373592c51e4b8. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B"}. This record does not assert independent reproduction. ### Average Exact Match (EM): 90. Result ID: 6a3ce002fbc373592c51e4c5. [Reporting paper]() Recorded evaluation configuration: {"Method\_Group":"3-Agent"}. This record does not assert independent reproduction. ### 3-Agent Exact Match (EM): 89.8. Result ID: 6a3ce002fbc373592c51e4b6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B"}. This record does not assert independent reproduction. ### Average Exact Match (EM): 89.7. Result ID: 6a3ce002fbc373592c51e4c4. [Reporting paper]() Recorded evaluation configuration: {"Method\_Group":"NF"}. This record does not assert independent reproduction. ### NF Exact Match (EM): 89.6. Result ID: 6a3ce002fbc373592c51e4b5. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B"}. This record does not assert independent reproduction. ### CoT Exact Match (EM): 89.5. Result ID: 6a3ce002fbc373592c51e4b7. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-8B"}. This record does not assert independent reproduction. ### 3-Agent Exact Match (EM): 88.9. Result ID: 6a3ce002fbc373592c51e4bb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### NF Exact Match (EM): 88.7. Result ID: 6a3ce002fbc373592c51e4ab. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### CoT Exact Match (EM): 88.4. Result ID: 6a3ce002fbc373592c51e4ad. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### SDA Exact Match (EM): 88.2. Result ID: 6a3ce002fbc373592c51e4ae. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### MADARA Exact Match (EM): 88.2. Result ID: 6a3ce002fbc373592c51e4be. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B","RSC":"Quality-Ordered","ρ̂1":"0.62","Strategy":"ATF†"}. This record does not assert independent reproduction. ### 3-Agent Exact Match (EM): 87.8. Result ID: 6a3ce002fbc373592c51e4ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### CoT Exact Match (EM): 87.5. Result ID: 6a3ce002fbc373592c51e4bc. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### NF Exact Match (EM): 87.1. Result ID: 6a3ce002fbc373592c51e4ba. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.