{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:JRBUWS5AGJXTP4RHNZOTMAMEQP","short_pith_number":"pith:JRBUWS5A","canonical_record":{"source":{"id":"2606.04928","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-03T14:21:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6f45ae0a1aafece67619f89cc5876ce33cb1d4342804481955b41a3cbb24a5cb","abstract_canon_sha256":"a3a8b73954f0e77b0a18fd2aeb8d1c071d1a87f21bc1a9e1f200f9d4f2fda0a7"},"schema_version":"1.0"},"canonical_sha256":"4c434b4ba0326f37f2276e5d36018483c474b22b7b0930a930c567252cbfc643","source":{"kind":"arxiv","id":"2606.04928","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.04928","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"arxiv_version","alias_value":"2606.04928v1","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.04928","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"pith_short_12","alias_value":"JRBUWS5AGJXT","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"pith_short_16","alias_value":"JRBUWS5AGJXTP4RH","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"pith_short_8","alias_value":"JRBUWS5A","created_at":"2026-06-04T01:09:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:JRBUWS5AGJXTP4RHNZOTMAMEQP","target":"record","payload":{"canonical_record":{"source":{"id":"2606.04928","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-03T14:21:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6f45ae0a1aafece67619f89cc5876ce33cb1d4342804481955b41a3cbb24a5cb","abstract_canon_sha256":"a3a8b73954f0e77b0a18fd2aeb8d1c071d1a87f21bc1a9e1f200f9d4f2fda0a7"},"schema_version":"1.0"},"canonical_sha256":"4c434b4ba0326f37f2276e5d36018483c474b22b7b0930a930c567252cbfc643","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-04T01:09:56.163802Z","signature_b64":"PQRs1taKxiYR9zRPs4u1DdjbvED2ApBAO+RlHsZ3CP1hDwryFHboleFIUh/VeFE7H70cE+Lo15Cn/8x8L2NbDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4c434b4ba0326f37f2276e5d36018483c474b22b7b0930a930c567252cbfc643","last_reissued_at":"2026-06-04T01:09:56.162971Z","signature_status":"signed_v1","first_computed_at":"2026-06-04T01:09:56.162971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.04928","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-04T01:09:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OSkwCRwkUA8PmdLWxTTUEYLf0nhEo5ijgi9nPdLxvFjLYaaR8gtsJhy9uVRaeWL9/tB9TBAf+vlt80Dqv1whDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T16:30:44.487013Z"},"content_sha256":"b2615868a3df1ed5fa408ae7b6e416260a1f43db83991a3fc2d383adf55c1a90","schema_version":"1.0","event_id":"sha256:b2615868a3df1ed5fa408ae7b6e416260a1f43db83991a3fc2d383adf55c1a90"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:JRBUWS5AGJXTP4RHNZOTMAMEQP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Data Attribution in Large Language Models via Bidirectional Gradient Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Fr\\'ed\\'eric Berdoz, Kaan Bayraktar, Luca A. Lanzend\\\"orfer, Roger Wattenhofer","submitted_at":"2026-06-03T14:21:53Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly deployed across diverse applications, raising critical questions for governance, accountability, and data provenance. Understanding which training data most influenced a model's output remains a fundamental open problem. We address this challenge through training data attribution (TDA) for auto-regressive LLMs by expanding upon the inverse formulation: How would training data be affected if the model had seen the generated output during training? Our method perturbs the base model using bidirectional gradient optimization (gradient ascent and desce"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.04928","kind":"arxiv","version":1},"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/2606.04928/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":"2026-06-04T01:09:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NfdBscWTQwnsytOuuTat7J7pcxz6Sy+SPVIN2jAxgY4C3VjWAFDAEhbR8mnI5Dnol5HzoalgjNIcoNiRVw6vDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T16:30:44.487770Z"},"content_sha256":"4c937893b896ce6b425d55bc64f318bf8bc07c39adf323f16192c2d75d2247a0","schema_version":"1.0","event_id":"sha256:4c937893b896ce6b425d55bc64f318bf8bc07c39adf323f16192c2d75d2247a0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP/bundle.json","state_url":"https://pith.science/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP/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":"2026-06-06T16:30:44Z","links":{"resolver":"https://pith.science/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP","bundle":"https://pith.science/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP/bundle.json","state":"https://pith.science/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JRBUWS5AGJXTP4RHNZOTMAMEQP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:JRBUWS5AGJXTP4RHNZOTMAMEQP","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":"a3a8b73954f0e77b0a18fd2aeb8d1c071d1a87f21bc1a9e1f200f9d4f2fda0a7","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-03T14:21:53Z","title_canon_sha256":"6f45ae0a1aafece67619f89cc5876ce33cb1d4342804481955b41a3cbb24a5cb"},"schema_version":"1.0","source":{"id":"2606.04928","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.04928","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"arxiv_version","alias_value":"2606.04928v1","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.04928","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"pith_short_12","alias_value":"JRBUWS5AGJXT","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"pith_short_16","alias_value":"JRBUWS5AGJXTP4RH","created_at":"2026-06-04T01:09:56Z"},{"alias_kind":"pith_short_8","alias_value":"JRBUWS5A","created_at":"2026-06-04T01:09:56Z"}],"graph_snapshots":[{"event_id":"sha256:4c937893b896ce6b425d55bc64f318bf8bc07c39adf323f16192c2d75d2247a0","target":"graph","created_at":"2026-06-04T01:09:56Z","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/2606.04928/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are increasingly deployed across diverse applications, raising critical questions for governance, accountability, and data provenance. Understanding which training data most influenced a model's output remains a fundamental open problem. We address this challenge through training data attribution (TDA) for auto-regressive LLMs by expanding upon the inverse formulation: How would training data be affected if the model had seen the generated output during training? Our method perturbs the base model using bidirectional gradient optimization (gradient ascent and desce","authors_text":"Fr\\'ed\\'eric Berdoz, Kaan Bayraktar, Luca A. Lanzend\\\"orfer, Roger Wattenhofer","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-03T14:21:53Z","title":"Data Attribution in Large Language Models via Bidirectional Gradient Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.04928","kind":"arxiv","version":1},"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:b2615868a3df1ed5fa408ae7b6e416260a1f43db83991a3fc2d383adf55c1a90","target":"record","created_at":"2026-06-04T01:09:56Z","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":"a3a8b73954f0e77b0a18fd2aeb8d1c071d1a87f21bc1a9e1f200f9d4f2fda0a7","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-03T14:21:53Z","title_canon_sha256":"6f45ae0a1aafece67619f89cc5876ce33cb1d4342804481955b41a3cbb24a5cb"},"schema_version":"1.0","source":{"id":"2606.04928","kind":"arxiv","version":1}},"canonical_sha256":"4c434b4ba0326f37f2276e5d36018483c474b22b7b0930a930c567252cbfc643","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4c434b4ba0326f37f2276e5d36018483c474b22b7b0930a930c567252cbfc643","first_computed_at":"2026-06-04T01:09:56.162971Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-04T01:09:56.162971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"PQRs1taKxiYR9zRPs4u1DdjbvED2ApBAO+RlHsZ3CP1hDwryFHboleFIUh/VeFE7H70cE+Lo15Cn/8x8L2NbDA==","signature_status":"signed_v1","signed_at":"2026-06-04T01:09:56.163802Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.04928","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b2615868a3df1ed5fa408ae7b6e416260a1f43db83991a3fc2d383adf55c1a90","sha256:4c937893b896ce6b425d55bc64f318bf8bc07c39adf323f16192c2d75d2247a0"],"state_sha256":"779eb3a93d4e6056fb5a3801168868212748c0dc9c88f635d2d432dafa7d12a1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"a/jBi+eV2GELZRy8GAqrrfUKr91z7xAIdyfn5V/cg77ro5ghOP+OZnNxB0eyos7m7a3V8+8kquY8Uz2mJ87QAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-06T16:30:44.490450Z","bundle_sha256":"f3335bae7295238b5dcb4175977d6e731161422f895c3c7b51f2d5393d477543"}}