# Circuit Localization on Mixing dataset IOI Source: [Circuit Localization on Mixing dataset IOI]() Collection index: [llms.txt]() Data updated: 2026-05-12T11:21:30.905Z Evaluation of model performance on the Circuit Localization task using the Mixing dataset IOI split, reporting the CMD metric. 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 | Circuit Localization | | Dataset (including any stated split) | Mixing dataset IOI | | Metric | CMD | | Unit | SCALAR | | Direction | Lower 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 - [CMD](): Lower is better; unit: SCALAR. ## Recorded results | Method | CMD | Result date | Report paper | | --- | --- | --- | --- | | K-Random | 0.008 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-kmeans-PCA | 0.008 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-kmeans-SVD | 0.009 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-agglom.-SVD | 0.011 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-divisive-SVD | 0.011 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-kmeans-PCA | 0.012 | Not recorded | 6a02b45cdbc836b87d8249e8 | | EAP-IG | 0.015 | Not recorded | 6a02b45cdbc836b87d8249e8 | | K-Random | 0.022 | Not recorded | 6a02b45cdbc836b87d8249e8 | | K-Random | 0.022 | Not recorded | 6a02b45cdbc836b87d8249e8 | | K-representative | 0.022 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-kmeans-SVD | 0.022 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-kmeans-PCA | 0.022 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-kmeans-SVD | 0.025 | Not recorded | 6a02b45cdbc836b87d8249e8 | | EAP | 0.026 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-divisive-SVD | 0.026 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-divisive-SVD | 0.026 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-agglom.-SVD | 0.047 | Not recorded | 6a02b45cdbc836b87d8249e8 | | DCD-agglom.-SVD | 0.047 | Not recorded | 6a02b45cdbc836b87d8249e8 | | EAP | 0.063 | Not recorded | 6a02b45cdbc836b87d8249e8 | | EAP | 0.068 | Not recorded | 6a02b45cdbc836b87d8249e8 | | K-representative | 0.071 | Not recorded | 6a02b45cdbc836b87d8249e8 | | K-representative | 0.071 | Not recorded | 6a02b45cdbc836b87d8249e8 | | EAP-IG | 0.125 | Not recorded | 6a02b45cdbc836b87d8249e8 | | EAP-IG | 0.125 | Not recorded | 6a02b45cdbc836b87d8249e8 | | E-Act | 0.129 | Not recorded | 6a02b45cdbc836b87d8249e8 | | Random | 0.749 | Not recorded | 6a02b45cdbc836b87d8249e8 | | Random | 0.845 | Not recorded | 6a02b45cdbc836b87d8249e8 | | Random | 0.845 | Not recorded | 6a02b45cdbc836b87d8249e8 | ### K-Random CMD: 0.008. Result ID: 6a030c7d47f34ce924e6cc0a. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-kmeans-PCA CMD: 0.008. Result ID: 6a030c7d47f34ce924e6cc19. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-kmeans-SVD CMD: 0.009. Result ID: 6a030c7d47f34ce924e6cc16. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-agglom.-SVD CMD: 0.011. Result ID: 6a030c7d47f34ce924e6cc10. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-divisive-SVD CMD: 0.011. Result ID: 6a030c7d47f34ce924e6cc13. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-kmeans-PCA CMD: 0.012. Result ID: 6a030c7d47f34ce924e6cc1a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### EAP-IG CMD: 0.015. Result ID: 6a030c7d47f34ce924e6cc00. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### K-Random CMD: 0.022. Result ID: 6a030c7d47f34ce924e6cc09. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### K-Random CMD: 0.022. Result ID: 6a030c7d47f34ce924e6cc0b. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### K-representative CMD: 0.022. Result ID: 6a030c7d47f34ce924e6cc0d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-kmeans-SVD CMD: 0.022. Result ID: 6a030c7d47f34ce924e6cc17. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### DCD-kmeans-PCA CMD: 0.022. Result ID: 6a030c7d47f34ce924e6cc18. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### DCD-kmeans-SVD CMD: 0.025. Result ID: 6a030c7d47f34ce924e6cc15. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### EAP CMD: 0.026. Result ID: 6a030c7d47f34ce924e6cc03. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### DCD-divisive-SVD CMD: 0.026. Result ID: 6a030c7d47f34ce924e6cc12. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### DCD-divisive-SVD CMD: 0.026. Result ID: 6a030c7d47f34ce924e6cc14. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### DCD-agglom.-SVD CMD: 0.047. Result ID: 6a030c7d47f34ce924e6cc0f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### DCD-agglom.-SVD CMD: 0.047. Result ID: 6a030c7d47f34ce924e6cc11. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### EAP CMD: 0.063. Result ID: 6a030c7d47f34ce924e6cc04. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### EAP CMD: 0.068. Result ID: 6a030c7d47f34ce924e6cc02. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### K-representative CMD: 0.071. Result ID: 6a030c7d47f34ce924e6cc0c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### K-representative CMD: 0.071. Result ID: 6a030c7d47f34ce924e6cc0e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### EAP-IG CMD: 0.125. Result ID: 6a030c7d47f34ce924e6cbff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### EAP-IG CMD: 0.125. Result ID: 6a030c7d47f34ce924e6cc01. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. ### E-Act CMD: 0.129. Result ID: 6a030c7d47f34ce924e6cc05. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### Random CMD: 0.749. Result ID: 6a030c7d47f34ce924e6cc07. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama"}. This record does not assert independent reproduction. ### Random CMD: 0.845. Result ID: 6a030c7d47f34ce924e6cc06. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT2"}. This record does not assert independent reproduction. ### Random CMD: 0.845. Result ID: 6a030c7d47f34ce924e6cc08. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.