{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "ameisen-et-al-circuit-tracing-2025", "title": "Ameisen and colleagues: cross-layer-transcoder circuit tracing", "summary": "Built prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks.", "boundary": "Attention patterns and normalization are frozen; reconstruction errors, missing attention circuits and incomplete mechanisms limit explanations. Graphs are not complete or globally faithful model proofs.", "area": "Mechanistic interpretability", "topic": "Artificial intelligence", "year": 2025, "lane": "HUMAN", "result": "ALGORITHM", "package_id": "ameisen-et-al-circuit-tracing-2025", "target": { "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "content_hash": "sha256:ddeb129c07f3f8167781a9e43e3b5a94a4ca6b3639b513106cdc11865571ba3e" }, "humans": [ "Emmanuel Ameisen", "Jack Lindsey", "Adam Pearce", "Wes Gurnee", "Nicholas L. Turner", "Brian Chen", "Craig Citro", "David Abrahams", "Shan Carter", "Basil Hosmer", "Jonathan Marcus", "Michael Sklar", "Adly Templeton", "Trenton Bricken", "Callum McDougall", "Hoagy Cunningham", "Thomas Henighan", "Adam Jermyn", "Andy Jones", "Andrew Persic", "Zhenyi Qi", "T. Ben Thompson", "Sam Zimmerman", "Kelley Rivoire", "Thomas Conerly", "Chris Olah", "Joshua Batson" ], "authors_complete": true, "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.", "agents": [], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "cur20261015-anthropic-circuit-tracing", "title": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } ], "points": "120000", "published_grade": { "id": "ameisen-et-al-circuit-tracing-2025", "entry_id": "ameisen-et-al-circuit-tracing-2025", "band": "A", "modifier": null, "target": { "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "content_hash": "sha256:ddeb129c07f3f8167781a9e43e3b5a94a4ca6b3639b513106cdc11865571ba3e" }, "what_changed": "Built prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks.", "scope": "Built prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks. Attention patterns and normalization are frozen; reconstruction errors, missing attention circuits and incomplete mechanisms limit explanations. Graphs are not complete or globally faithful model proofs.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "content_hash": "sha256:ddeb129c07f3f8167781a9e43e3b5a94a4ca6b3639b513106cdc11865571ba3e" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-10-16T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [ { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Emmanuel Ameisen", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jack Lindsey", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Adam Pearce", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Wes Gurnee", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Nicholas L. Turner", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Brian Chen", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Craig Citro", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "David Abrahams", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Shan Carter", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Basil Hosmer", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jonathan Marcus", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Michael Sklar", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Adly Templeton", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Trenton Bricken", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Callum McDougall", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Hoagy Cunningham", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Thomas Henighan", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Adam Jermyn", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Andy Jones", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Andrew Persic", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zhenyi Qi", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "T. Ben Thompson", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Sam Zimmerman", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Kelley Rivoire", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Thomas Conerly", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Chris Olah", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Joshua Batson", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "Attention patterns and normalization are frozen; reconstruction errors, missing attention circuits and incomplete mechanisms limit explanations. Graphs are not complete or globally faithful model proofs.", "domain": "Mechanistic interpretability" }, "sources": [ { "citation": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html" } ], "statement": "# Ameisen and colleagues: cross-layer-transcoder circuit tracing\n\nBuilt prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks.\n\n## Scientific scope\n\nAttention patterns and normalization are frozen; reconstruction errors, missing attention circuits and incomplete mechanisms limit explanations. Graphs are not complete or globally faithful model proofs.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nEmmanuel Ameisen; Jack Lindsey; Adam Pearce; Wes Gurnee; Nicholas L. Turner; Brian Chen; Craig Citro; David Abrahams; Shan Carter; Basil Hosmer; Jonathan Marcus; Michael Sklar; Adly Templeton; Trenton Bricken; Callum McDougall; Hoagy Cunningham; Thomas Henighan; Adam Jermyn; Andy Jones; Andrew Persic; Zhenyi Qi; T. Ben Thompson; Sam Zimmerman; Kelley Rivoire; Thomas Conerly; Chris Olah; Joshua Batson\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [Circuit Tracing: Revealing Computational Graphs in Language Models](https://transformer-circuits.pub/2025/attribution-graphs/methods.html)\n", "summary": "Built prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks.", "title": "Ameisen and colleagues: cross-layer-transcoder circuit tracing" }, "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "source_ref": { "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "content_hash": "sha256:ddeb129c07f3f8167781a9e43e3b5a94a4ca6b3639b513106cdc11865571ba3e" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:bcecfb20ac343ee1321160201d428154e89e9841e962cab3f8543e069f5b0d5e" }, "target": { "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "content_hash": "sha256:ddeb129c07f3f8167781a9e43e3b5a94a4ca6b3639b513106cdc11865571ba3e" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "ameisen-et-al-circuit-tracing-2025", "entry_id": "ameisen-et-al-circuit-tracing-2025", "revision": 1, "title": "Ameisen and colleagues: cross-layer-transcoder circuit tracing", "context": "Mechanistic interpretability", "track": "Human-designed research method or resource", "band": "A", "modifier": null, "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-10-16", "evidence_cutoff": "2026-10-16", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "Built prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks.", "scope": "Built prompt-specific attribution graphs using cross-layer transcoders as an interpretable replacement for model MLP computation, with intervention and faithfulness checks. Attention patterns and normalization are frozen; reconstruction errors, missing attention circuits and incomplete mechanisms limit explanations. Graphs are not complete or globally faithful model proofs.", "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Attention patterns and normalization are frozen; reconstruction errors, missing attention circuits and incomplete mechanisms limit explanations. Graphs are not complete or globally faithful model proofs.", "references": [ { "title": "Circuit Tracing: Revealing Computational Graphs in Language Models", "url": "https://transformer-circuits.pub/2025/attribution-graphs/methods.html", "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification." } ], "roles": [ { "who": "Source-listed authors and technical contributors", "credit": "SCIENCE", "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities." } ], "target": { "record_id": "urn:uuid:640eb97e-354e-46f7-a754-1e2d0053440c", "version": 1, "content_hash": "sha256:ddeb129c07f3f8167781a9e43e3b5a94a4ca6b3639b513106cdc11865571ba3e" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "120000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": null, "reading_status": "SOURCE_LINKED_NOTE" }