{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:OOWKI5UAOH5MKEBICLBNRLXT6N","short_pith_number":"pith:OOWKI5UA","canonical_record":{"source":{"id":"2504.13151","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-04-17T17:55:45Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"4745586e5cb98cb022e5f8593d3c7d33da0524d03fb414d0ba72294b6a9b33c8","abstract_canon_sha256":"f235d6a2d66725f3016b9637f0f63d19ab5e9a88d135012c932296b57ab42c99"},"schema_version":"1.0"},"canonical_sha256":"73aca4768071fac5102812c2d8aef3f3568140700cbe4838bebfcf990d5eb5fa","source":{"kind":"arxiv","id":"2504.13151","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.13151","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"arxiv_version","alias_value":"2504.13151v2","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.13151","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"pith_short_12","alias_value":"OOWKI5UAOH5M","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"pith_short_16","alias_value":"OOWKI5UAOH5MKEBI","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"pith_short_8","alias_value":"OOWKI5UA","created_at":"2026-07-05T11:18:20Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:OOWKI5UAOH5MKEBICLBNRLXT6N","target":"record","payload":{"canonical_record":{"source":{"id":"2504.13151","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-04-17T17:55:45Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"4745586e5cb98cb022e5f8593d3c7d33da0524d03fb414d0ba72294b6a9b33c8","abstract_canon_sha256":"f235d6a2d66725f3016b9637f0f63d19ab5e9a88d135012c932296b57ab42c99"},"schema_version":"1.0"},"canonical_sha256":"73aca4768071fac5102812c2d8aef3f3568140700cbe4838bebfcf990d5eb5fa","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:20.359509Z","signature_b64":"5Fs2nHBWBUaGxbaIiRYmlLshWEZAGdZ9cBWtCyclbhP9Kw+3HJXKUFbZGINwUSJHxqW5muQat3Gd+RXOfWFbDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"73aca4768071fac5102812c2d8aef3f3568140700cbe4838bebfcf990d5eb5fa","last_reissued_at":"2026-07-05T11:18:20.359020Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:20.359020Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.13151","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:18:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"q+b8U5xrTObKmwF15WK8PkN1Fmf4jYef0rkv9ZkKoTjMCAeAqb9tvCjEpsOK9/D3RCHtQYUOkwWLwLpXZxfxCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T15:58:52.611396Z"},"content_sha256":"d2293795e4cdb92c8eead7f7de57cd1fc1ffa151cf623f9708018a619ecfb590","schema_version":"1.0","event_id":"sha256:d2293795e4cdb92c8eead7f7de57cd1fc1ffa151cf623f9708018a619ecfb590"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:OOWKI5UAOH5MKEBICLBNRLXT6N","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MIB: A Mechanistic Interpretability Benchmark","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Aaron Mueller, Adam Belfki, Alessandro Stolfo, Amir Zur, Anja Reusch, Aruna Sankaranarayanan, Atticus Geiger, Dana Arad, David Bau, Hadas Orgad, Iv\\'an Arcuschin, Jaden Fiotto-Kaufman, Jing Huang, Martin Tutek, Michael Hanna, Nikhil Prakash, Rohan Gupta, Sarah Wiegreffe, Shun Shao, Tal Haklay, Yaniv Nikankin, Yik Siu Chan, Yonatan Belinkov","submitted_at":"2025-04-17T17:55:45Z","abstract_excerpt":"How can we know whether new mechanistic interpretability methods achieve real improvements? In pursuit of lasting evaluation standards, we propose MIB, a Mechanistic Interpretability Benchmark, with two tracks spanning four tasks and five models. MIB favors methods that precisely and concisely recover relevant causal pathways or causal variables in neural language models. The circuit localization track compares methods that locate the model components - and connections between them - most important for performing a task (e.g., attribution patching or information flow routes). The causal variab"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.13151","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2504.13151/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:18:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qPGeBQKNqjROCHJIRdItAIC3vc1eKsfiqXV6OZuj6mOxohfwyRQ8tKhNYWDs+EUef8P2VFVv/LDrudTTk1qYCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T15:58:52.611924Z"},"content_sha256":"98e371281aff6ce3f77737d7e6eca20b89a3ad96ed96bb471a0524a2f7072dd5","schema_version":"1.0","event_id":"sha256:98e371281aff6ce3f77737d7e6eca20b89a3ad96ed96bb471a0524a2f7072dd5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OOWKI5UAOH5MKEBICLBNRLXT6N/bundle.json","state_url":"https://pith.science/pith/OOWKI5UAOH5MKEBICLBNRLXT6N/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OOWKI5UAOH5MKEBICLBNRLXT6N/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-15T15:58:52Z","links":{"resolver":"https://pith.science/pith/OOWKI5UAOH5MKEBICLBNRLXT6N","bundle":"https://pith.science/pith/OOWKI5UAOH5MKEBICLBNRLXT6N/bundle.json","state":"https://pith.science/pith/OOWKI5UAOH5MKEBICLBNRLXT6N/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OOWKI5UAOH5MKEBICLBNRLXT6N/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:OOWKI5UAOH5MKEBICLBNRLXT6N","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"f235d6a2d66725f3016b9637f0f63d19ab5e9a88d135012c932296b57ab42c99","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-04-17T17:55:45Z","title_canon_sha256":"4745586e5cb98cb022e5f8593d3c7d33da0524d03fb414d0ba72294b6a9b33c8"},"schema_version":"1.0","source":{"id":"2504.13151","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.13151","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"arxiv_version","alias_value":"2504.13151v2","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.13151","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"pith_short_12","alias_value":"OOWKI5UAOH5M","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"pith_short_16","alias_value":"OOWKI5UAOH5MKEBI","created_at":"2026-07-05T11:18:20Z"},{"alias_kind":"pith_short_8","alias_value":"OOWKI5UA","created_at":"2026-07-05T11:18:20Z"}],"graph_snapshots":[{"event_id":"sha256:98e371281aff6ce3f77737d7e6eca20b89a3ad96ed96bb471a0524a2f7072dd5","target":"graph","created_at":"2026-07-05T11:18:20Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2504.13151/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"How can we know whether new mechanistic interpretability methods achieve real improvements? In pursuit of lasting evaluation standards, we propose MIB, a Mechanistic Interpretability Benchmark, with two tracks spanning four tasks and five models. MIB favors methods that precisely and concisely recover relevant causal pathways or causal variables in neural language models. The circuit localization track compares methods that locate the model components - and connections between them - most important for performing a task (e.g., attribution patching or information flow routes). The causal variab","authors_text":"Aaron Mueller, Adam Belfki, Alessandro Stolfo, Amir Zur, Anja Reusch, Aruna Sankaranarayanan, Atticus Geiger, Dana Arad, David Bau, Hadas Orgad, Iv\\'an Arcuschin, Jaden Fiotto-Kaufman, Jing Huang, Martin Tutek, Michael Hanna, Nikhil Prakash, Rohan Gupta, Sarah Wiegreffe, Shun Shao, Tal Haklay, Yaniv Nikankin, Yik Siu Chan, Yonatan Belinkov","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-04-17T17:55:45Z","title":"MIB: A Mechanistic Interpretability Benchmark"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.13151","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d2293795e4cdb92c8eead7f7de57cd1fc1ffa151cf623f9708018a619ecfb590","target":"record","created_at":"2026-07-05T11:18:20Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"f235d6a2d66725f3016b9637f0f63d19ab5e9a88d135012c932296b57ab42c99","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-04-17T17:55:45Z","title_canon_sha256":"4745586e5cb98cb022e5f8593d3c7d33da0524d03fb414d0ba72294b6a9b33c8"},"schema_version":"1.0","source":{"id":"2504.13151","kind":"arxiv","version":2}},"canonical_sha256":"73aca4768071fac5102812c2d8aef3f3568140700cbe4838bebfcf990d5eb5fa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"73aca4768071fac5102812c2d8aef3f3568140700cbe4838bebfcf990d5eb5fa","first_computed_at":"2026-07-05T11:18:20.359020Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:18:20.359020Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5Fs2nHBWBUaGxbaIiRYmlLshWEZAGdZ9cBWtCyclbhP9Kw+3HJXKUFbZGINwUSJHxqW5muQat3Gd+RXOfWFbDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:18:20.359509Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.13151","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d2293795e4cdb92c8eead7f7de57cd1fc1ffa151cf623f9708018a619ecfb590","sha256:98e371281aff6ce3f77737d7e6eca20b89a3ad96ed96bb471a0524a2f7072dd5"],"state_sha256":"a327988614e31d0b12c454df40298380beade39f67da008b2dd1fa56b8c88498"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mhUcKPOT9aVFyligPRBeIdQYm1Kmgq2wrizd7wh1zW8yQD5YVVs2oFQEk0oX+XHuHxxCampGRNTs+FCuLUnZBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T15:58:52.617145Z","bundle_sha256":"b59f935394448b84d089ff97b349aa7291d7b7a01b28faf11fd858968c310ff3"}}