# Causal Reasoning on e-CARE OOD (held-out) Source: [Causal Reasoning on e-CARE OOD (held-out)]() Collection index: [llms.txt]() Data updated: 2026-06-16T08:44:06.625Z Evaluation of model performance on the held-out Out-of-Distribution (OOD) subset of the e-CARE dataset for Causal Reasoning tasks, measured by P0, P1, Gap, and Relative Gap. 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 | Causal Reasoning | | Dataset (including any stated split) | e-CARE OOD (held-out) | | Metric | P0 Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 13 of 13 | | 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 - [P0 Score](): Higher is better; unit: PERCENT. - [P1 Score](): Higher is better; unit: PERCENT. - [Causal Gap](): Higher is better; unit: SCALAR. - [Relative Gap](): Lower is better; unit: PERCENT. ## Recorded results | Method | P0 Score | Result date | Report paper | | --- | --- | --- | --- | | Qwen2.5-14B | 82.3 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Qwen2.5-14B | 82 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Qwen2.5-32B | 82 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Qwen2.5-7B | 81.5 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Qwen2.5-32B | 81.5 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Qwen2.5-7B | 81 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Phi-3-mini (3.8B) | 79.5 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Llama-3.1-8B | 79.3 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Llama-3.1-8B | 79 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Mistral-7B | 79 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Phi-3-mini (3.8B) | 78.5 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Mistral-7B | 77.5 | Not recorded | 6a30d89fd33dfb23e783cc86 | | Mistral-7B | 77 | Not recorded | 6a30d89fd33dfb23e783cc86 | ### Qwen2.5-14B P0 Score: 82.3. Result ID: 6a310cc3fbc373592c2d67f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier","QLoRA":"true"}. This record does not assert independent reproduction. ### Qwen2.5-14B P0 Score: 82. Result ID: 6a310cc3fbc373592c2d67f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Base","QLoRA":"true"}. This record does not assert independent reproduction. ### Qwen2.5-32B P0 Score: 82. Result ID: 6a310cc3fbc373592c2d67f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Base","QLoRA":"true"}. This record does not assert independent reproduction. ### Qwen2.5-7B P0 Score: 81.5. Result ID: 6a310cc3fbc373592c2d67f0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Base"}. This record does not assert independent reproduction. ### Qwen2.5-32B P0 Score: 81.5. Result ID: 6a310cc3fbc373592c2d67f5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier","QLoRA":"true"}. This record does not assert independent reproduction. ### Qwen2.5-7B P0 Score: 81. Result ID: 6a310cc3fbc373592c2d67f1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier"}. This record does not assert independent reproduction. ### Phi-3-mini (3.8B) P0 Score: 79.5. Result ID: 6a310cc3fbc373592c2d67ee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Base"}. This record does not assert independent reproduction. ### Llama-3.1-8B P0 Score: 79.3. Result ID: 6a310cc3fbc373592c2d67f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier"}. This record does not assert independent reproduction. ### Llama-3.1-8B P0 Score: 79. Result ID: 6a310cc3fbc373592c2d67f6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Base"}. This record does not assert independent reproduction. ### Mistral-7B P0 Score: 79. Result ID: 6a310cc3fbc373592c2d67f8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Base"}. This record does not assert independent reproduction. ### Phi-3-mini (3.8B) P0 Score: 78.5. Result ID: 6a310cc3fbc373592c2d67ef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier"}. This record does not assert independent reproduction. ### Mistral-7B P0 Score: 77.5. Result ID: 6a310cc3fbc373592c2d67f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier","recipe":"default","lr":"1e-4"}. This record does not assert independent reproduction. ### Mistral-7B P0 Score: 77. Result ID: 6a310cc3fbc373592c2d67fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Adapter type":"Vernier","recipe":"rescued","lr":"5e-6","early stopping step":"150"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.