{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/localizing-model-behavior-with-path-patching","title":"Localizing Model Behavior with Path Patching","arxiv_id":"2304.05969","date":"2023-04-12","proceeding":null,"authors":["Nicholas Goldowsky-Dill","Chris MacLeod","Lucas Sato","Aryaman Arora"],"abstract":"Localizing behaviors of neural networks to a subset of the network's components or a subset of interactions between components is a natural first step towards analyzing network mechanisms and possible failure modes. Existing work is often qualitative and ad-hoc, and there is no consensus on the appropriate way to evaluate localization claims. We introduce path patching, a technique for expressing and quantitatively testing a natural class of hypotheses expressing that behaviors are localized to a set of paths. We refine an explanation of induction heads, characterize a behavior of GPT-2, and open source a framework for efficiently running similar experiments.","url_abs":"https://arxiv.org/abs/2304.05969v2","url_pdf":"https://arxiv.org/pdf/2304.05969v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"localizing-model-behavior-with-path-patching","repo_url":"https://github.com/redwoodresearch/rust_circuit_public","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"jax","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"model","task_name":"model"}],"methods":[{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"attention-dropout","method_name":"Attention Dropout"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"cosine-annealing","method_name":"Cosine Annealing"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"discriminative-fine-tuning","method_name":"Discriminative Fine-Tuning"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"gpt-2","method_name":"GPT-2"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"linear-warmup-with-cosine-annealing","method_name":"Linear Warmup With Cosine Annealing"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"weight-decay","method_name":"Weight Decay"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2304.05969","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2304.05969"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/redwoodresearch/rust_circuit_public","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":13},"by_repo_kind":{"official":{"samples":13,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"45d0115e5f739bf8","entry":"check_cast","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/py_utils.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/py_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"45d0115e5f739bf8"}},{"code_sha256_prefix":"1a02c7bfaa2b8388","entry":"cum_name_to_eps_name","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/cum_algo.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/cum_algo.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1a02c7bfaa2b8388"}},{"code_sha256_prefix":"39872ad6e0f9c6ca","entry":"dedup_with_order","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/algebric_rewrite.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/algebric_rewrite.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"39872ad6e0f9c6ca"}},{"code_sha256_prefix":"383f1fbd21776cf4","entry":"from_converted","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/jax_to_module.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/jax_to_module.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"383f1fbd21776cf4"}},{"code_sha256_prefix":"d3d17b52b27c0023","entry":"from_converted_tup","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/jax_to_module.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/jax_to_module.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d3d17b52b27c0023"}},{"code_sha256_prefix":"bd83c07b355d6df4","entry":"get_tensor_shape","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"rr_util/rust_circuit_type_utils.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/rr_util/rust_circuit_type_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"bd83c07b355d6df4"}},{"code_sha256_prefix":"8f38ed9a551bc6bc","entry":"map","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/optional.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/optional.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8f38ed9a551bc6bc"}},{"code_sha256_prefix":"b8768128ba49da6e","entry":"maybe_dtype_to_maybe_string","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"rr_util/rust_circuit_type_utils.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/rr_util/rust_circuit_type_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b8768128ba49da6e"}},{"code_sha256_prefix":"35ca6e3dd2e5742c","entry":"normalize_index","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/algebric_rewrite.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/algebric_rewrite.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"35ca6e3dd2e5742c"}},{"code_sha256_prefix":"ea9e811ba6f76498","entry":"py_generalfunction_get_shape","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/interop_rust.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/interop_rust.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ea9e811ba6f76498"}},{"code_sha256_prefix":"32014b3f5ce2b2b7","entry":"scalar_to_tensor","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"rr_util/rust_circuit_type_utils.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/rr_util/rust_circuit_type_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"32014b3f5ce2b2b7"}},{"code_sha256_prefix":"4cab52cbaf461422","entry":"unwrap","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/optional.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/optional.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4cab52cbaf461422"}},{"code_sha256_prefix":"0b29322d64014266","entry":"unwrap_or","repo":"redwoodresearch/rust_circuit_public","repo_kind":"official","path":"python/rust_circuit/optional.py","file_url":"https://github.com/redwoodresearch/rust_circuit_public/blob/HEAD/python/rust_circuit/optional.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0b29322d64014266"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}