{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:UKKCQNEX4EFGEB7DQAUJ7E64YW","short_pith_number":"pith:UKKCQNEX","canonical_record":{"source":{"id":"2605.20730","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-20T05:26:38Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6f9f02845399ecf0c43e70f3f565170b9a07549701e5d828cdb82ebb8e1babc4","abstract_canon_sha256":"9d4cedcc4819f93bc9a76fefb1abee8e5371f60fbdaa432bc1a3ad6ebcdd6811"},"schema_version":"1.0"},"canonical_sha256":"a294283497e10a6207e380289f93dcc59b9a0d7c89d58a3b4b23ca2cf433a01f","source":{"kind":"arxiv","id":"2605.20730","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.20730","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"arxiv_version","alias_value":"2605.20730v1","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20730","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"pith_short_12","alias_value":"UKKCQNEX4EFG","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"pith_short_16","alias_value":"UKKCQNEX4EFGEB7D","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"pith_short_8","alias_value":"UKKCQNEX","created_at":"2026-05-21T01:04:51Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:UKKCQNEX4EFGEB7DQAUJ7E64YW","target":"record","payload":{"canonical_record":{"source":{"id":"2605.20730","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-20T05:26:38Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6f9f02845399ecf0c43e70f3f565170b9a07549701e5d828cdb82ebb8e1babc4","abstract_canon_sha256":"9d4cedcc4819f93bc9a76fefb1abee8e5371f60fbdaa432bc1a3ad6ebcdd6811"},"schema_version":"1.0"},"canonical_sha256":"a294283497e10a6207e380289f93dcc59b9a0d7c89d58a3b4b23ca2cf433a01f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:51.195671Z","signature_b64":"8NeIF3e9H+bl026YKquRctVajn3i5d2nOqUZkyJLscnJ4pxthQEeM1sbhOLfL0N59duzuKNRxPQn7oj1A1N6CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a294283497e10a6207e380289f93dcc59b9a0d7c89d58a3b4b23ca2cf433a01f","last_reissued_at":"2026-05-21T01:04:51.194941Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:51.194941Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.20730","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":"2026-05-21T01:04:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"arZ0P/uNaXVDZBO7fsXJuCcCc+kTea0XFfHrhqnB8IwddV0nN7Q9eqpkMcQXN67ATvHqMiu4thWEYUxXC41QAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T23:08:11.175158Z"},"content_sha256":"d668be014637a7721747c3b2af1434a19213a4b72289d32f8af0e486577eb3fc","schema_version":"1.0","event_id":"sha256:d668be014637a7721747c3b2af1434a19213a4b72289d32f8af0e486577eb3fc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:UKKCQNEX4EFGEB7DQAUJ7E64YW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Distributional Alignment as a Criterion for Designing Task Vectors in In-Context Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jihoon Kwon, Jiwon Choi, Jy-yong Sohn","submitted_at":"2026-05-20T05:26:38Z","abstract_excerpt":"In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks through demonstrations, yet it suffers from escalating inference costs as context length increases. While task vectors offer a promising alternative by compressing demonstrations into compact hidden-state representations, their quality has been evaluated only through downstream task accuracy. This indirect criterion provides limited insight into how to design more effective task vector extraction methods. In this paper, we posit that inference using task vectors should align their predictive distribution with t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20730","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/2605.20730/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":"2026-05-21T01:04:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fxcmeVKgak0moYtTb41Y50iDdjtBu6mOfmU2VArEf5M8SW39S8i2u2NMhQlQhFUFvrB5OblaqF6W4IBqWNEnDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T23:08:11.175946Z"},"content_sha256":"509d3384e5b7fae5d650e960fe11555e093f62bcf7c0eb6115f775116645dbe6","schema_version":"1.0","event_id":"sha256:509d3384e5b7fae5d650e960fe11555e093f62bcf7c0eb6115f775116645dbe6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW/bundle.json","state_url":"https://pith.science/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW/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":"2026-06-05T23:08:11Z","links":{"resolver":"https://pith.science/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW","bundle":"https://pith.science/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW/bundle.json","state":"https://pith.science/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UKKCQNEX4EFGEB7DQAUJ7E64YW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:UKKCQNEX4EFGEB7DQAUJ7E64YW","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":"9d4cedcc4819f93bc9a76fefb1abee8e5371f60fbdaa432bc1a3ad6ebcdd6811","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-20T05:26:38Z","title_canon_sha256":"6f9f02845399ecf0c43e70f3f565170b9a07549701e5d828cdb82ebb8e1babc4"},"schema_version":"1.0","source":{"id":"2605.20730","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.20730","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"arxiv_version","alias_value":"2605.20730v1","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20730","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"pith_short_12","alias_value":"UKKCQNEX4EFG","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"pith_short_16","alias_value":"UKKCQNEX4EFGEB7D","created_at":"2026-05-21T01:04:51Z"},{"alias_kind":"pith_short_8","alias_value":"UKKCQNEX","created_at":"2026-05-21T01:04:51Z"}],"graph_snapshots":[{"event_id":"sha256:509d3384e5b7fae5d650e960fe11555e093f62bcf7c0eb6115f775116645dbe6","target":"graph","created_at":"2026-05-21T01:04:51Z","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/2605.20730/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks through demonstrations, yet it suffers from escalating inference costs as context length increases. While task vectors offer a promising alternative by compressing demonstrations into compact hidden-state representations, their quality has been evaluated only through downstream task accuracy. This indirect criterion provides limited insight into how to design more effective task vector extraction methods. In this paper, we posit that inference using task vectors should align their predictive distribution with t","authors_text":"Jihoon Kwon, Jiwon Choi, Jy-yong Sohn","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-20T05:26:38Z","title":"Distributional Alignment as a Criterion for Designing Task Vectors in In-Context Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20730","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:d668be014637a7721747c3b2af1434a19213a4b72289d32f8af0e486577eb3fc","target":"record","created_at":"2026-05-21T01:04:51Z","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":"9d4cedcc4819f93bc9a76fefb1abee8e5371f60fbdaa432bc1a3ad6ebcdd6811","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-20T05:26:38Z","title_canon_sha256":"6f9f02845399ecf0c43e70f3f565170b9a07549701e5d828cdb82ebb8e1babc4"},"schema_version":"1.0","source":{"id":"2605.20730","kind":"arxiv","version":1}},"canonical_sha256":"a294283497e10a6207e380289f93dcc59b9a0d7c89d58a3b4b23ca2cf433a01f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a294283497e10a6207e380289f93dcc59b9a0d7c89d58a3b4b23ca2cf433a01f","first_computed_at":"2026-05-21T01:04:51.194941Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-21T01:04:51.194941Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8NeIF3e9H+bl026YKquRctVajn3i5d2nOqUZkyJLscnJ4pxthQEeM1sbhOLfL0N59duzuKNRxPQn7oj1A1N6CA==","signature_status":"signed_v1","signed_at":"2026-05-21T01:04:51.195671Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.20730","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d668be014637a7721747c3b2af1434a19213a4b72289d32f8af0e486577eb3fc","sha256:509d3384e5b7fae5d650e960fe11555e093f62bcf7c0eb6115f775116645dbe6"],"state_sha256":"0b19c1ffd8e7290c4bdf68c198aadeb225610f00d1aa8100ac1d1287c54e686f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lm5DQyTt9ZlSCREOxyWoWNnzIDxM0l2Gpuk9c5VfuoRZ1atCeOn0/2T6G8bvaVmDm4LIQAgwI/T2gmQhzPUPCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T23:08:11.180780Z","bundle_sha256":"f8eb6423a256793f9d3bbe6dda354ad46eedaf6cf5c956448515439a537ef426"}}