{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PFJMW7US64USOFDI7BNQSFBPK7","short_pith_number":"pith:PFJMW7US","schema_version":"1.0","canonical_sha256":"7952cb7e92f729271468f85b09142f57c24f4a309d4a8c2df4c786d7e834a4ad","source":{"kind":"arxiv","id":"2511.17864","version":3},"attestation_state":"computed","paper":{"title":"Equivalence of Context and Parameter Updates in Modern Transformer Blocks","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adrian Goldwaser, Benoit Dherin, Javier Gonzalvo, Michael Munn","submitted_at":"2025-11-22T01:17:15Z","abstract_excerpt":"Recent research has established that the impact of context in a vanilla transformer can be represented implicitly by forming a token-dependent, rank-1 patch to its MLP weights. This work extends that foundational theory to the diverse architectures of modern Large Language Models. We first demonstrate a precise, analytical solution for a Gemma-style transformer block, proving that the entire effect of a context can be perfectly mapped to rank-1 patches on its MLP weight matrices and a patch to the RMSNorm scale. We then generalize this result, providing a constructive proof and algorithm for m"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2511.17864","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-22T01:17:15Z","cross_cats_sorted":[],"title_canon_sha256":"e7575b2a57d3216a1df3f91188bf73f7229a6480a0f65cd905161812b01a73a2","abstract_canon_sha256":"2ceabf34ab0f9858a5bb1cd278c9b04bf2d19283fc6399fcad51d23ee1ac6cfb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T03:18:50.640733Z","signature_b64":"lAxiNqvuX1tTVFH74EoNH0c1Me1ILlZG9fxzbQg/eysQ5tw3h7xbSDfZYFbdx10CvhwAeQiaIhgyDWm7r8z6CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7952cb7e92f729271468f85b09142f57c24f4a309d4a8c2df4c786d7e834a4ad","last_reissued_at":"2026-07-07T03:18:50.640167Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T03:18:50.640167Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Equivalence of Context and Parameter Updates in Modern Transformer Blocks","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adrian Goldwaser, Benoit Dherin, Javier Gonzalvo, Michael Munn","submitted_at":"2025-11-22T01:17:15Z","abstract_excerpt":"Recent research has established that the impact of context in a vanilla transformer can be represented implicitly by forming a token-dependent, rank-1 patch to its MLP weights. This work extends that foundational theory to the diverse architectures of modern Large Language Models. We first demonstrate a precise, analytical solution for a Gemma-style transformer block, proving that the entire effect of a context can be perfectly mapped to rank-1 patches on its MLP weight matrices and a patch to the RMSNorm scale. We then generalize this result, providing a constructive proof and algorithm for m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.17864","kind":"arxiv","version":3},"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/2511.17864/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2511.17864","created_at":"2026-07-07T03:18:50.640227+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.17864v3","created_at":"2026-07-07T03:18:50.640227+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.17864","created_at":"2026-07-07T03:18:50.640227+00:00"},{"alias_kind":"pith_short_12","alias_value":"PFJMW7US64US","created_at":"2026-07-07T03:18:50.640227+00:00"},{"alias_kind":"pith_short_16","alias_value":"PFJMW7US64USOFDI","created_at":"2026-07-07T03:18:50.640227+00:00"},{"alias_kind":"pith_short_8","alias_value":"PFJMW7US","created_at":"2026-07-07T03:18:50.640227+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.32022","citing_title":"SemRF: A Semantic Reference Frame for Residual-Stream Dynamics in Language Models","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7","json":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7.json","graph_json":"https://pith.science/api/pith-number/PFJMW7US64USOFDI7BNQSFBPK7/graph.json","events_json":"https://pith.science/api/pith-number/PFJMW7US64USOFDI7BNQSFBPK7/events.json","paper":"https://pith.science/paper/PFJMW7US"},"agent_actions":{"view_html":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7","download_json":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7.json","view_paper":"https://pith.science/paper/PFJMW7US","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.17864&json=true","fetch_graph":"https://pith.science/api/pith-number/PFJMW7US64USOFDI7BNQSFBPK7/graph.json","fetch_events":"https://pith.science/api/pith-number/PFJMW7US64USOFDI7BNQSFBPK7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7/action/storage_attestation","attest_author":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7/action/author_attestation","sign_citation":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7/action/citation_signature","submit_replication":"https://pith.science/pith/PFJMW7US64USOFDI7BNQSFBPK7/action/replication_record"}},"created_at":"2026-07-07T03:18:50.640227+00:00","updated_at":"2026-07-07T03:18:50.640227+00:00"}