{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:PVWAVXLRURM45KC4LU4VKQJV74","short_pith_number":"pith:PVWAVXLR","canonical_record":{"source":{"id":"2409.13710","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T16:17:06Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ac292f47eb14172372f7c3e04c40edb3af6f050c4ff47d94cb25dc81aff54c06","abstract_canon_sha256":"22bd7113140f3ecb78577e035cf240892535776de4be05b339de8416a1586e6b"},"schema_version":"1.0"},"canonical_sha256":"7d6c0add71a459cea85c5d39554135ff288ec722d9e6272711bdbb0c40cd9613","source":{"kind":"arxiv","id":"2409.13710","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.13710","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"arxiv_version","alias_value":"2409.13710v2","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.13710","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"pith_short_12","alias_value":"PVWAVXLRURM4","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"pith_short_16","alias_value":"PVWAVXLRURM45KC4","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"pith_short_8","alias_value":"PVWAVXLR","created_at":"2026-07-05T09:36:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:PVWAVXLRURM45KC4LU4VKQJV74","target":"record","payload":{"canonical_record":{"source":{"id":"2409.13710","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T16:17:06Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ac292f47eb14172372f7c3e04c40edb3af6f050c4ff47d94cb25dc81aff54c06","abstract_canon_sha256":"22bd7113140f3ecb78577e035cf240892535776de4be05b339de8416a1586e6b"},"schema_version":"1.0"},"canonical_sha256":"7d6c0add71a459cea85c5d39554135ff288ec722d9e6272711bdbb0c40cd9613","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:36:48.230083Z","signature_b64":"VTzWifVKeu0YeJLACoVdEjqyD3XnZYF5OrOBzr8nMrH87WZgCn9cxALW47KM/orDwSvVDgjLUFoWbiDVcOwYAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d6c0add71a459cea85c5d39554135ff288ec722d9e6272711bdbb0c40cd9613","last_reissued_at":"2026-07-05T09:36:48.229607Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:36:48.229607Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.13710","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":"2026-07-05T09:36:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6p6H8pkjmY2IRDvRAbtjyjr0SX73tIl8lIak/lSWB1AM9m3GjR0WIq7KvaY5wQVk3LTagRQvBjQow6H8TNMuAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T11:25:05.631863Z"},"content_sha256":"d31d46c612827de8c79f5e7e69c088acf50133e7590d844668c7348a5968192c","schema_version":"1.0","event_id":"sha256:d31d46c612827de8c79f5e7e69c088acf50133e7590d844668c7348a5968192c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:PVWAVXLRURM45KC4LU4VKQJV74","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"You can remove GPT2's LayerNorm by fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Stefan Heimersheim","submitted_at":"2024-09-06T16:17:06Z","abstract_excerpt":"The LayerNorm (LN) layer in GPT-style transformer models has long been a hindrance to mechanistic interpretability. LN is a crucial component required to stabilize the training of large language models, and LN or the similar RMSNorm have been used in practically all large language models based on the transformer architecture. The non-linear nature of the LN layers is a hindrance for mechanistic interpretability as it hinders interpretation of the residual stream, and makes it difficult to decompose the model into circuits. Some researchers have gone so far as to name \"reasons interpretability "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.13710","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/2409.13710/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":"2026-07-05T09:36:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dCgpjozwkkrcvubY9BEY7hHeOVWoLT6OldC44ObrC950grlGf02boxtZrBttYITUHOzytgVUnpd/Z15VgEAGBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T11:25:05.632749Z"},"content_sha256":"09c5b7065f4031c88482148b03a4622d26c84e40b7c24174b5927f2f39fe3536","schema_version":"1.0","event_id":"sha256:09c5b7065f4031c88482148b03a4622d26c84e40b7c24174b5927f2f39fe3536"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PVWAVXLRURM45KC4LU4VKQJV74/bundle.json","state_url":"https://pith.science/pith/PVWAVXLRURM45KC4LU4VKQJV74/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PVWAVXLRURM45KC4LU4VKQJV74/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":"2026-08-09T11:25:05Z","links":{"resolver":"https://pith.science/pith/PVWAVXLRURM45KC4LU4VKQJV74","bundle":"https://pith.science/pith/PVWAVXLRURM45KC4LU4VKQJV74/bundle.json","state":"https://pith.science/pith/PVWAVXLRURM45KC4LU4VKQJV74/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PVWAVXLRURM45KC4LU4VKQJV74/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:PVWAVXLRURM45KC4LU4VKQJV74","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":"22bd7113140f3ecb78577e035cf240892535776de4be05b339de8416a1586e6b","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T16:17:06Z","title_canon_sha256":"ac292f47eb14172372f7c3e04c40edb3af6f050c4ff47d94cb25dc81aff54c06"},"schema_version":"1.0","source":{"id":"2409.13710","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.13710","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"arxiv_version","alias_value":"2409.13710v2","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.13710","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"pith_short_12","alias_value":"PVWAVXLRURM4","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"pith_short_16","alias_value":"PVWAVXLRURM45KC4","created_at":"2026-07-05T09:36:48Z"},{"alias_kind":"pith_short_8","alias_value":"PVWAVXLR","created_at":"2026-07-05T09:36:48Z"}],"graph_snapshots":[{"event_id":"sha256:09c5b7065f4031c88482148b03a4622d26c84e40b7c24174b5927f2f39fe3536","target":"graph","created_at":"2026-07-05T09:36:48Z","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/2409.13710/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The LayerNorm (LN) layer in GPT-style transformer models has long been a hindrance to mechanistic interpretability. LN is a crucial component required to stabilize the training of large language models, and LN or the similar RMSNorm have been used in practically all large language models based on the transformer architecture. The non-linear nature of the LN layers is a hindrance for mechanistic interpretability as it hinders interpretation of the residual stream, and makes it difficult to decompose the model into circuits. Some researchers have gone so far as to name \"reasons interpretability ","authors_text":"Stefan Heimersheim","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T16:17:06Z","title":"You can remove GPT2's LayerNorm by fine-tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.13710","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:d31d46c612827de8c79f5e7e69c088acf50133e7590d844668c7348a5968192c","target":"record","created_at":"2026-07-05T09:36:48Z","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":"22bd7113140f3ecb78577e035cf240892535776de4be05b339de8416a1586e6b","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-09-06T16:17:06Z","title_canon_sha256":"ac292f47eb14172372f7c3e04c40edb3af6f050c4ff47d94cb25dc81aff54c06"},"schema_version":"1.0","source":{"id":"2409.13710","kind":"arxiv","version":2}},"canonical_sha256":"7d6c0add71a459cea85c5d39554135ff288ec722d9e6272711bdbb0c40cd9613","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7d6c0add71a459cea85c5d39554135ff288ec722d9e6272711bdbb0c40cd9613","first_computed_at":"2026-07-05T09:36:48.229607Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:36:48.229607Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VTzWifVKeu0YeJLACoVdEjqyD3XnZYF5OrOBzr8nMrH87WZgCn9cxALW47KM/orDwSvVDgjLUFoWbiDVcOwYAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:36:48.230083Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.13710","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d31d46c612827de8c79f5e7e69c088acf50133e7590d844668c7348a5968192c","sha256:09c5b7065f4031c88482148b03a4622d26c84e40b7c24174b5927f2f39fe3536"],"state_sha256":"3e4d416ce41b0a2914c5bfb01fb686af5291c1e6c846d65e07ae467a2df7dec4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Lr7PXstrja2/Kjgg8rGXg8JHD2NukZvx85R0zAEaeFM2iJbH04it7JBTZmbjmLCmrlZ6BffnefazxvglMPWNAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T11:25:05.640547Z","bundle_sha256":"a559d3bf72799416c1f8cc2784cb831b04e53a233005c3a4f66a15b734ae7483"}}