{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:NNEVGTUC6DHTDXXM7HGSEHDZIG","short_pith_number":"pith:NNEVGTUC","canonical_record":{"source":{"id":"2502.04602","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-07T01:32:19Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"51e8bd60468c7f629efc2b0f91d4a94f17b63b90fc0e81d902f1f1abb0c13000","abstract_canon_sha256":"53183a2a52001effadf8aba6ffe0252f10889f3a6bd14c0add649976aa91271a"},"schema_version":"1.0"},"canonical_sha256":"6b49534e82f0cf31deecf9cd221c7941affe1dc2ada5d869258aac5d8ae05301","source":{"kind":"arxiv","id":"2502.04602","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.04602","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"arxiv_version","alias_value":"2502.04602v1","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04602","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"pith_short_12","alias_value":"NNEVGTUC6DHT","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"pith_short_16","alias_value":"NNEVGTUC6DHTDXXM","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"pith_short_8","alias_value":"NNEVGTUC","created_at":"2026-07-05T10:10:54Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:NNEVGTUC6DHTDXXM7HGSEHDZIG","target":"record","payload":{"canonical_record":{"source":{"id":"2502.04602","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-07T01:32:19Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"51e8bd60468c7f629efc2b0f91d4a94f17b63b90fc0e81d902f1f1abb0c13000","abstract_canon_sha256":"53183a2a52001effadf8aba6ffe0252f10889f3a6bd14c0add649976aa91271a"},"schema_version":"1.0"},"canonical_sha256":"6b49534e82f0cf31deecf9cd221c7941affe1dc2ada5d869258aac5d8ae05301","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:54.185373Z","signature_b64":"bP+E0CZyP6SxRVXnB7NZQmVDbeUgemJfvQFbcDXa7vr5vTgIoFiqBOSR+l+rAKkr8bLOxbLgu1Z/rD3NdUMNBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b49534e82f0cf31deecf9cd221c7941affe1dc2ada5d869258aac5d8ae05301","last_reissued_at":"2026-07-05T10:10:54.184801Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:54.184801Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.04602","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":"2026-07-05T10:10:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+2tL7ojSpCZaM4YhyN8CkkN5kBm1vRjhvH8RXJeisu0OduWtmxCWB79n9s6tPgKsHxkyds37wl3tlLLz1oYtAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T04:50:13.685047Z"},"content_sha256":"1e8619a70a0a43c0f7a3955549f8e972b8f3eb58090bbc31bf8e7caa18ab4482","schema_version":"1.0","event_id":"sha256:1e8619a70a0a43c0f7a3955549f8e972b8f3eb58090bbc31bf8e7caa18ab4482"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:NNEVGTUC6DHTDXXM7HGSEHDZIG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Extracting and Understanding the Superficial Knowledge in Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bhavya Kailkhura, Gabriel Jacob Perin, Junyuan Hong, Nina S. T. Hirata, Runjin Chen, Xilun Chen, Xuxi Chen, Yan Han","submitted_at":"2025-02-07T01:32:19Z","abstract_excerpt":"Alignment of large language models (LLMs) with human values and preferences, often achieved through fine-tuning based on human feedback, is essential for ensuring safe and responsible AI behaviors. However, the process typically requires substantial data and computation resources. Recent studies have revealed that alignment might be attainable at lower costs through simpler methods, such as in-context learning. This leads to the question: Is alignment predominantly superficial? In this paper, we delve into this question and provide a quantitative analysis. We formalize the concept of superfici"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04602","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/2502.04602/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":"2026-07-05T10:10:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DGU1yhlPxVNbPx9jqEI/VXsTkYSFgzoEoIlKDhnKvY6oi3xoqULmdKuLK9WFo8l/AavZAW1DcRhjvjvLY8geDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T04:50:13.685611Z"},"content_sha256":"265e406950017b2d8811d3d7accb77c1b6690958a4aee949a9b141bcd74700e6","schema_version":"1.0","event_id":"sha256:265e406950017b2d8811d3d7accb77c1b6690958a4aee949a9b141bcd74700e6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG/bundle.json","state_url":"https://pith.science/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG/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":"2026-08-09T04:50:13Z","links":{"resolver":"https://pith.science/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG","bundle":"https://pith.science/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG/bundle.json","state":"https://pith.science/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NNEVGTUC6DHTDXXM7HGSEHDZIG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:NNEVGTUC6DHTDXXM7HGSEHDZIG","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":"53183a2a52001effadf8aba6ffe0252f10889f3a6bd14c0add649976aa91271a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-07T01:32:19Z","title_canon_sha256":"51e8bd60468c7f629efc2b0f91d4a94f17b63b90fc0e81d902f1f1abb0c13000"},"schema_version":"1.0","source":{"id":"2502.04602","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.04602","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"arxiv_version","alias_value":"2502.04602v1","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04602","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"pith_short_12","alias_value":"NNEVGTUC6DHT","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"pith_short_16","alias_value":"NNEVGTUC6DHTDXXM","created_at":"2026-07-05T10:10:54Z"},{"alias_kind":"pith_short_8","alias_value":"NNEVGTUC","created_at":"2026-07-05T10:10:54Z"}],"graph_snapshots":[{"event_id":"sha256:265e406950017b2d8811d3d7accb77c1b6690958a4aee949a9b141bcd74700e6","target":"graph","created_at":"2026-07-05T10:10:54Z","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/2502.04602/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Alignment of large language models (LLMs) with human values and preferences, often achieved through fine-tuning based on human feedback, is essential for ensuring safe and responsible AI behaviors. However, the process typically requires substantial data and computation resources. Recent studies have revealed that alignment might be attainable at lower costs through simpler methods, such as in-context learning. This leads to the question: Is alignment predominantly superficial? In this paper, we delve into this question and provide a quantitative analysis. We formalize the concept of superfici","authors_text":"Bhavya Kailkhura, Gabriel Jacob Perin, Junyuan Hong, Nina S. T. Hirata, Runjin Chen, Xilun Chen, Xuxi Chen, Yan Han","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-07T01:32:19Z","title":"Extracting and Understanding the Superficial Knowledge in Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04602","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":"sha256:1e8619a70a0a43c0f7a3955549f8e972b8f3eb58090bbc31bf8e7caa18ab4482","target":"record","created_at":"2026-07-05T10:10:54Z","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":"53183a2a52001effadf8aba6ffe0252f10889f3a6bd14c0add649976aa91271a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-07T01:32:19Z","title_canon_sha256":"51e8bd60468c7f629efc2b0f91d4a94f17b63b90fc0e81d902f1f1abb0c13000"},"schema_version":"1.0","source":{"id":"2502.04602","kind":"arxiv","version":1}},"canonical_sha256":"6b49534e82f0cf31deecf9cd221c7941affe1dc2ada5d869258aac5d8ae05301","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6b49534e82f0cf31deecf9cd221c7941affe1dc2ada5d869258aac5d8ae05301","first_computed_at":"2026-07-05T10:10:54.184801Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:10:54.184801Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"bP+E0CZyP6SxRVXnB7NZQmVDbeUgemJfvQFbcDXa7vr5vTgIoFiqBOSR+l+rAKkr8bLOxbLgu1Z/rD3NdUMNBg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:10:54.185373Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.04602","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1e8619a70a0a43c0f7a3955549f8e972b8f3eb58090bbc31bf8e7caa18ab4482","sha256:265e406950017b2d8811d3d7accb77c1b6690958a4aee949a9b141bcd74700e6"],"state_sha256":"56b0aa12b946249936cd8e659594542dd1382c4e63a9424268d370af46422035"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0ex7+/GuypGY0ivycPtFeUVCkwKFuYKZJn6bbcXCw94sNkDAfTzPsLJirVZRmXEjBBonIH6KHvGbALqubnUjBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T04:50:13.689449Z","bundle_sha256":"4f6c5046b84e2b1ec42753451a612a16dbe9d9481f1115db4f83c2d107e121d4"}}