# Causal Faithfulness Evaluation on Edgar Domain Source: [Causal Faithfulness Evaluation on Edgar Domain]() Collection index: [llms.txt]() Data updated: 2026-05-22T07:41:28.538Z Evaluation of model causal faithfulness on the Edgar Domain dataset, measured using nAOPC, nComp, and \|∆CE\| 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 | Causal Faithfulness Evaluation | | Dataset (including any stated split) | Edgar Domain | | Metric | nAOPC | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 28 of 28 | | 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 - [nAOPC](): Higher is better; unit: SCALAR. - [nComp](): Higher is better; unit: SCALAR. - [\|∆CE\|](): Lower is better; unit: SCALAR. ## Recorded results | Method | nAOPC | Result date | Report paper | | --- | --- | --- | --- | | GAE | 0.988 | Not recorded | 6a0fe320f984739ed6739e05 | | GAE | 0.981 | Not recorded | 6a0fe320f984739ed6739e05 | | SAEBoost | 0.979 | Not recorded | 6a0fe320f984739ed6739e05 | | FaithfulSAE | 0.976 | Not recorded | 6a0fe320f984739ed6739e05 | | Fixed | 0.975 | Not recorded | 6a0fe320f984739ed6739e05 | | Finetune | 0.971 | Not recorded | 6a0fe320f984739ed6739e05 | | Retrain | 0.97 | Not recorded | 6a0fe320f984739ed6739e05 | | GAE | 0.968 | Not recorded | 6a0fe320f984739ed6739e05 | | TERM | 0.964 | Not recorded | 6a0fe320f984739ed6739e05 | | TERM | 0.959 | Not recorded | 6a0fe320f984739ed6739e05 | | Finetune | 0.959 | Not recorded | 6a0fe320f984739ed6739e05 | | SAEBoost | 0.955 | Not recorded | 6a0fe320f984739ed6739e05 | | Fixed | 0.953 | Not recorded | 6a0fe320f984739ed6739e05 | | FaithfulSAE | 0.953 | Not recorded | 6a0fe320f984739ed6739e05 | | Retrain | 0.936 | Not recorded | 6a0fe320f984739ed6739e05 | | SAEBoost | 0.903 | Not recorded | 6a0fe320f984739ed6739e05 | | TERM | 0.895 | Not recorded | 6a0fe320f984739ed6739e05 | | Retrain | 0.746 | Not recorded | 6a0fe320f984739ed6739e05 | | Fixed | 0.725 | Not recorded | 6a0fe320f984739ed6739e05 | | FaithfulSAE | 0.724 | Not recorded | 6a0fe320f984739ed6739e05 | | GAE | 0.723 | Not recorded | 6a0fe320f984739ed6739e05 | | Retrain | 0.715 | Not recorded | 6a0fe320f984739ed6739e05 | | Finetune | 0.684 | Not recorded | 6a0fe320f984739ed6739e05 | | Finetune | 0.658 | Not recorded | 6a0fe320f984739ed6739e05 | | FaithfulSAE | 0.657 | Not recorded | 6a0fe320f984739ed6739e05 | | TERM | 0.655 | Not recorded | 6a0fe320f984739ed6739e05 | | Fixed | 0.65 | Not recorded | 6a0fe320f984739ed6739e05 | | SAEBoost | 0.604 | Not recorded | 6a0fe320f984739ed6739e05 | ### GAE nAOPC: 0.988. Result ID: 6a1008a4fbc373592c08afb2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### GAE nAOPC: 0.981. Result ID: 6a1008a4fbc373592c08afab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### SAEBoost nAOPC: 0.979. Result ID: 6a1008a4fbc373592c08afa9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### FaithfulSAE nAOPC: 0.976. Result ID: 6a1008a4fbc373592c08afaa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### Fixed nAOPC: 0.975. Result ID: 6a1008a4fbc373592c08afa5. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### Finetune nAOPC: 0.971. Result ID: 6a1008a4fbc373592c08afa7. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### Retrain nAOPC: 0.97. Result ID: 6a1008a4fbc373592c08afd9. [Reporting paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### GAE nAOPC: 0.968. Result ID: 6a1008a4fbc373592c08afdc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### TERM nAOPC: 0.964. Result ID: 6a1008a4fbc373592c08afa6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### TERM nAOPC: 0.959. Result ID: 6a1008a4fbc373592c08afd7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### Finetune nAOPC: 0.959. Result ID: 6a1008a4fbc373592c08afd8. [Reporting paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### SAEBoost nAOPC: 0.955. Result ID: 6a1008a4fbc373592c08afda. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### Fixed nAOPC: 0.953. Result ID: 6a1008a4fbc373592c08afd6. [Reporting paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### FaithfulSAE nAOPC: 0.953. Result ID: 6a1008a4fbc373592c08afdb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### Retrain nAOPC: 0.936. Result ID: 6a1008a4fbc373592c08afa8. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### SAEBoost nAOPC: 0.903. Result ID: 6a1008a4fbc373592c08afb0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### TERM nAOPC: 0.895. Result ID: 6a1008a4fbc373592c08afad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### Retrain nAOPC: 0.746. Result ID: 6a1008a4fbc373592c08afaf. [Reporting paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### Fixed nAOPC: 0.725. Result ID: 6a1008a4fbc373592c08afac. [Reporting paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### FaithfulSAE nAOPC: 0.724. Result ID: 6a1008a4fbc373592c08afb1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### GAE nAOPC: 0.723. Result ID: 6a1008a4fbc373592c08afd5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### Retrain nAOPC: 0.715. Result ID: 6a1008a4fbc373592c08afd2. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### Finetune nAOPC: 0.684. Result ID: 6a1008a4fbc373592c08afae. [Reporting paper]() Recorded evaluation configuration: {"Model":"Pythia-1.4B"}. This record does not assert independent reproduction. ### Finetune nAOPC: 0.658. Result ID: 6a1008a4fbc373592c08afd1. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### FaithfulSAE nAOPC: 0.657. Result ID: 6a1008a4fbc373592c08afd4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### TERM nAOPC: 0.655. Result ID: 6a1008a4fbc373592c08afd0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### Fixed nAOPC: 0.65. Result ID: 6a1008a4fbc373592c08afcf. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. ### SAEBoost nAOPC: 0.604. Result ID: 6a1008a4fbc373592c08afd3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-2 Small"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.