{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:AJEG4UMESLLJTAPSHODU63KEOZ","short_pith_number":"pith:AJEG4UME","canonical_record":{"source":{"id":"2606.06320","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-04T15:56:32Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"c3831da6b877d823e1a66f1cc7a85aac2a0365c97ef35a26fc21a91146e50824","abstract_canon_sha256":"dcd51f5f0e471a474c51cb8e2fee165bb6b917c50007f4910eb67537f5e9a537"},"schema_version":"1.0"},"canonical_sha256":"02486e518492d69981f23b874f6d447662da172719ae80eb43570660bb35cdfc","source":{"kind":"arxiv","id":"2606.06320","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.06320","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"arxiv_version","alias_value":"2606.06320v1","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.06320","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"pith_short_12","alias_value":"AJEG4UMESLLJ","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"pith_short_16","alias_value":"AJEG4UMESLLJTAPS","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"pith_short_8","alias_value":"AJEG4UME","created_at":"2026-06-05T01:15:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:AJEG4UMESLLJTAPSHODU63KEOZ","target":"record","payload":{"canonical_record":{"source":{"id":"2606.06320","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-04T15:56:32Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"c3831da6b877d823e1a66f1cc7a85aac2a0365c97ef35a26fc21a91146e50824","abstract_canon_sha256":"dcd51f5f0e471a474c51cb8e2fee165bb6b917c50007f4910eb67537f5e9a537"},"schema_version":"1.0"},"canonical_sha256":"02486e518492d69981f23b874f6d447662da172719ae80eb43570660bb35cdfc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:15:41.865505Z","signature_b64":"eXpOM5NWBtd5AyqSIlXnNDTB4j0oaAPjInJRK9iYVczt3b58zPKsT18j8VpCMBA8x6JP8tK06rSxXHgggKx3DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"02486e518492d69981f23b874f6d447662da172719ae80eb43570660bb35cdfc","last_reissued_at":"2026-06-05T01:15:41.865094Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:15:41.865094Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.06320","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":"2026-06-05T01:15:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GVkyVFSPnCAaym9SMIPDDkk19pLwY2PRMq2UP/1MV2XF9FuOKtxocIn3Kg7WOZ0iRhEJHdFCaehKM19P4cyNDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T15:22:47.469757Z"},"content_sha256":"4e1ecb6f9e352c1699eb2d1a91df137903121f21adbfb98bc8e1e891e213ab31","schema_version":"1.0","event_id":"sha256:4e1ecb6f9e352c1699eb2d1a91df137903121f21adbfb98bc8e1e891e213ab31"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:AJEG4UMESLLJTAPSHODU63KEOZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Learning What to Forget: Improving LLM Unlearning via Learned Token-Level Importance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Giorgos Nikolaou, Gizem Y\\\"uce, Nicolas Flammarion","submitted_at":"2026-06-04T15:56:32Z","abstract_excerpt":"Machine unlearning aims to remove targeted knowledge from a trained model while preserving its general capabilities. For autoregressive language models, not all tokens in a forget sample are equally relevant to forgetting. Existing approaches either ignore this heterogeneity or rely on auxiliary models, heuristics, or external annotations to estimate each token's relevance for forgetting. We instead characterize it through the interaction with the retain objective: a token is forget-specific to the extent that minimizing the forget loss on that token does not conflict with retain optimality. W"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.06320","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/2606.06320/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":"2026-06-05T01:15:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UFKeVuoDmKqzpdJTz/tJdyAkOw4YfSz7Oyl+GL89Zp91pPRcjMWwRww/Vho8xfNokX+WXMMf09eIeIv2p0ycDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T15:22:47.470132Z"},"content_sha256":"780bc90129fee82a540372bd0921dc279b3653484165174341f695a24dde5765","schema_version":"1.0","event_id":"sha256:780bc90129fee82a540372bd0921dc279b3653484165174341f695a24dde5765"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AJEG4UMESLLJTAPSHODU63KEOZ/bundle.json","state_url":"https://pith.science/pith/AJEG4UMESLLJTAPSHODU63KEOZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AJEG4UMESLLJTAPSHODU63KEOZ/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":"2026-06-29T15:22:47Z","links":{"resolver":"https://pith.science/pith/AJEG4UMESLLJTAPSHODU63KEOZ","bundle":"https://pith.science/pith/AJEG4UMESLLJTAPSHODU63KEOZ/bundle.json","state":"https://pith.science/pith/AJEG4UMESLLJTAPSHODU63KEOZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AJEG4UMESLLJTAPSHODU63KEOZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:AJEG4UMESLLJTAPSHODU63KEOZ","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":"dcd51f5f0e471a474c51cb8e2fee165bb6b917c50007f4910eb67537f5e9a537","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-04T15:56:32Z","title_canon_sha256":"c3831da6b877d823e1a66f1cc7a85aac2a0365c97ef35a26fc21a91146e50824"},"schema_version":"1.0","source":{"id":"2606.06320","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.06320","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"arxiv_version","alias_value":"2606.06320v1","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.06320","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"pith_short_12","alias_value":"AJEG4UMESLLJ","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"pith_short_16","alias_value":"AJEG4UMESLLJTAPS","created_at":"2026-06-05T01:15:41Z"},{"alias_kind":"pith_short_8","alias_value":"AJEG4UME","created_at":"2026-06-05T01:15:41Z"}],"graph_snapshots":[{"event_id":"sha256:780bc90129fee82a540372bd0921dc279b3653484165174341f695a24dde5765","target":"graph","created_at":"2026-06-05T01:15:41Z","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/2606.06320/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Machine unlearning aims to remove targeted knowledge from a trained model while preserving its general capabilities. For autoregressive language models, not all tokens in a forget sample are equally relevant to forgetting. Existing approaches either ignore this heterogeneity or rely on auxiliary models, heuristics, or external annotations to estimate each token's relevance for forgetting. We instead characterize it through the interaction with the retain objective: a token is forget-specific to the extent that minimizing the forget loss on that token does not conflict with retain optimality. W","authors_text":"Giorgos Nikolaou, Gizem Y\\\"uce, Nicolas Flammarion","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-04T15:56:32Z","title":"Learning What to Forget: Improving LLM Unlearning via Learned Token-Level Importance"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.06320","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":"sha256:4e1ecb6f9e352c1699eb2d1a91df137903121f21adbfb98bc8e1e891e213ab31","target":"record","created_at":"2026-06-05T01:15:41Z","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":"dcd51f5f0e471a474c51cb8e2fee165bb6b917c50007f4910eb67537f5e9a537","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-04T15:56:32Z","title_canon_sha256":"c3831da6b877d823e1a66f1cc7a85aac2a0365c97ef35a26fc21a91146e50824"},"schema_version":"1.0","source":{"id":"2606.06320","kind":"arxiv","version":1}},"canonical_sha256":"02486e518492d69981f23b874f6d447662da172719ae80eb43570660bb35cdfc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"02486e518492d69981f23b874f6d447662da172719ae80eb43570660bb35cdfc","first_computed_at":"2026-06-05T01:15:41.865094Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-05T01:15:41.865094Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eXpOM5NWBtd5AyqSIlXnNDTB4j0oaAPjInJRK9iYVczt3b58zPKsT18j8VpCMBA8x6JP8tK06rSxXHgggKx3DA==","signature_status":"signed_v1","signed_at":"2026-06-05T01:15:41.865505Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.06320","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4e1ecb6f9e352c1699eb2d1a91df137903121f21adbfb98bc8e1e891e213ab31","sha256:780bc90129fee82a540372bd0921dc279b3653484165174341f695a24dde5765"],"state_sha256":"3478aaa81c0bc1b1336735bd7f2b694079f39f5ac25745ff6cc793b9d06b42f9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9CEus/tvCMwidCglysJImoQsFuYnG960Nst6td+9i8KS71z6ktcmm0+Ur6le5/KMAz7ZqtZ/KmVnYZKoZRCmCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T15:22:47.472090Z","bundle_sha256":"53d805db281909dbd2fec47e14b54725bd6acb8588eb5607d95cc696d30dfb6b"}}