{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RPRGYRJ7UIN7PSGUKYG4WJDVWX","short_pith_number":"pith:RPRGYRJ7","canonical_record":{"source":{"id":"2602.23197","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-26T16:49:15Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"7d6aa73533bbf7c6cc0c4031fd5a4c1baa760721d93474c6591d5cb242bf6897","abstract_canon_sha256":"9024a0b3d09223bff6187359e75dab36f3cbaae18ab767b2e2c99e9059c88e05"},"schema_version":"1.0"},"canonical_sha256":"8be26c453fa21bf7c8d4560dcb2475b5d71bb0055a4de99cab0c923552931b7b","source":{"kind":"arxiv","id":"2602.23197","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.23197","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"arxiv_version","alias_value":"2602.23197v2","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.23197","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"pith_short_12","alias_value":"RPRGYRJ7UIN7","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"pith_short_16","alias_value":"RPRGYRJ7UIN7PSGU","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"pith_short_8","alias_value":"RPRGYRJ7","created_at":"2026-06-02T02:04:15Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RPRGYRJ7UIN7PSGUKYG4WJDVWX","target":"record","payload":{"canonical_record":{"source":{"id":"2602.23197","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-26T16:49:15Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"7d6aa73533bbf7c6cc0c4031fd5a4c1baa760721d93474c6591d5cb242bf6897","abstract_canon_sha256":"9024a0b3d09223bff6187359e75dab36f3cbaae18ab767b2e2c99e9059c88e05"},"schema_version":"1.0"},"canonical_sha256":"8be26c453fa21bf7c8d4560dcb2475b5d71bb0055a4de99cab0c923552931b7b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:15.281954Z","signature_b64":"Ryqa7JlAVD7QRzZaiUJubATV3GGevlo1C5vn5SCrfT1sLeVJjNyd/KaJmFIOwehTwoTMwtVY5eOdcCkEfjLsAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8be26c453fa21bf7c8d4560dcb2475b5d71bb0055a4de99cab0c923552931b7b","last_reissued_at":"2026-06-02T02:04:15.281420Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:15.281420Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.23197","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":"2026-06-02T02:04:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"k8REyLHyWEDs1oaSn1ZNk/IBoJ9sm6eO8P2C3PCjj3MFeWv7CmYkhQVmb6nwrwZEx9RgIS97ciC5ClQ23t/hCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T21:57:04.960585Z"},"content_sha256":"0f985ad97035c4410a6f969d426f7b8993f60514c9473ee64b4c9f5b8b704aeb","schema_version":"1.0","event_id":"sha256:0f985ad97035c4410a6f969d426f7b8993f60514c9473ee64b4c9f5b8b704aeb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RPRGYRJ7UIN7PSGUKYG4WJDVWX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fine-Tuning Without Forgetting In-Context Learning: A Theoretical Analysis of Linear Attention Models","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Chungpa Lee, Jy-yong Sohn, Kangwook Lee","submitted_at":"2026-02-26T16:49:15Z","abstract_excerpt":"Transformer-based large language models exhibit in-context learning, enabling adaptation to downstream tasks via few-shot prompting with demonstrations. In practice, such models are often fine-tuned to improve zero-shot performance on downstream tasks, allowing them to solve tasks without examples and thereby reducing inference costs. However, fine-tuning can degrade in-context learning, limiting the performance of fine-tuned models on tasks not seen during fine-tuning. Using linear attention models, we provide a theoretical analysis that characterizes how fine-tuning objectives modify attenti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.23197","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/2602.23197/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":"2026-06-02T02:04:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IbzL7dF+rX+Zc2VHc10kXOcSm7fvlL1X3ZaE4QpG/6UpeYbr7UM7QVsBwFssp4UCUNnpI9/lL8rrx433gu7dAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T21:57:04.961177Z"},"content_sha256":"c36de4176fdaa797ef80c3d8063ce6ed3ee4209cdab53930f15130ee5c07a21b","schema_version":"1.0","event_id":"sha256:c36de4176fdaa797ef80c3d8063ce6ed3ee4209cdab53930f15130ee5c07a21b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX/bundle.json","state_url":"https://pith.science/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX/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":"2026-06-06T21:57:04Z","links":{"resolver":"https://pith.science/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX","bundle":"https://pith.science/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX/bundle.json","state":"https://pith.science/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RPRGYRJ7UIN7PSGUKYG4WJDVWX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RPRGYRJ7UIN7PSGUKYG4WJDVWX","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":"9024a0b3d09223bff6187359e75dab36f3cbaae18ab767b2e2c99e9059c88e05","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-26T16:49:15Z","title_canon_sha256":"7d6aa73533bbf7c6cc0c4031fd5a4c1baa760721d93474c6591d5cb242bf6897"},"schema_version":"1.0","source":{"id":"2602.23197","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.23197","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"arxiv_version","alias_value":"2602.23197v2","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.23197","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"pith_short_12","alias_value":"RPRGYRJ7UIN7","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"pith_short_16","alias_value":"RPRGYRJ7UIN7PSGU","created_at":"2026-06-02T02:04:15Z"},{"alias_kind":"pith_short_8","alias_value":"RPRGYRJ7","created_at":"2026-06-02T02:04:15Z"}],"graph_snapshots":[{"event_id":"sha256:c36de4176fdaa797ef80c3d8063ce6ed3ee4209cdab53930f15130ee5c07a21b","target":"graph","created_at":"2026-06-02T02:04:15Z","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/2602.23197/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Transformer-based large language models exhibit in-context learning, enabling adaptation to downstream tasks via few-shot prompting with demonstrations. In practice, such models are often fine-tuned to improve zero-shot performance on downstream tasks, allowing them to solve tasks without examples and thereby reducing inference costs. However, fine-tuning can degrade in-context learning, limiting the performance of fine-tuned models on tasks not seen during fine-tuning. Using linear attention models, we provide a theoretical analysis that characterizes how fine-tuning objectives modify attenti","authors_text":"Chungpa Lee, Jy-yong Sohn, Kangwook Lee","cross_cats":["cs.LG","stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-26T16:49:15Z","title":"Fine-Tuning Without Forgetting In-Context Learning: A Theoretical Analysis of Linear Attention Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.23197","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":"sha256:0f985ad97035c4410a6f969d426f7b8993f60514c9473ee64b4c9f5b8b704aeb","target":"record","created_at":"2026-06-02T02:04:15Z","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":"9024a0b3d09223bff6187359e75dab36f3cbaae18ab767b2e2c99e9059c88e05","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-26T16:49:15Z","title_canon_sha256":"7d6aa73533bbf7c6cc0c4031fd5a4c1baa760721d93474c6591d5cb242bf6897"},"schema_version":"1.0","source":{"id":"2602.23197","kind":"arxiv","version":2}},"canonical_sha256":"8be26c453fa21bf7c8d4560dcb2475b5d71bb0055a4de99cab0c923552931b7b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8be26c453fa21bf7c8d4560dcb2475b5d71bb0055a4de99cab0c923552931b7b","first_computed_at":"2026-06-02T02:04:15.281420Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:15.281420Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ryqa7JlAVD7QRzZaiUJubATV3GGevlo1C5vn5SCrfT1sLeVJjNyd/KaJmFIOwehTwoTMwtVY5eOdcCkEfjLsAQ==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:15.281954Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.23197","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0f985ad97035c4410a6f969d426f7b8993f60514c9473ee64b4c9f5b8b704aeb","sha256:c36de4176fdaa797ef80c3d8063ce6ed3ee4209cdab53930f15130ee5c07a21b"],"state_sha256":"2af549c035edee6f4a9712f5fcc70049b4ac0f36f6923863c76a6f6e4767ecf5"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iRiU0kiYS18E6PoSdzJzI8c76oX9hURbDhaKAC0v+a/AUs7KBAMfpKLPqM5VNH0B8CsDYnaRztaBM5HSzC1ZCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-06T21:57:04.964088Z","bundle_sha256":"8d9134059c2c7bc5a0b4300aa6390a81040c2dcf01713d610ff5b471acc941b9"}}