{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:AULZ3AZC67UNKGCRSKUH5RF5GP","short_pith_number":"pith:AULZ3AZC","canonical_record":{"source":{"id":"2602.15338","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-17T03:58:55Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"892ca8ca3244d3ec44dd8ab6b7cd228fef9aa52838a4c082d99c5a212b62e849","abstract_canon_sha256":"e36bae88821557c6b522dc193d2b16e86aa2ee353bdf41e23bf18704e5721a6e"},"schema_version":"1.0"},"canonical_sha256":"05179d8322f7e8d5185192a87ec4bd33d537203a2c3e3097dad012a3b01180fb","source":{"kind":"arxiv","id":"2602.15338","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.15338","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"arxiv_version","alias_value":"2602.15338v2","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.15338","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"pith_short_12","alias_value":"AULZ3AZC67UN","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"pith_short_16","alias_value":"AULZ3AZC67UNKGCR","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"pith_short_8","alias_value":"AULZ3AZC","created_at":"2026-05-22T01:03:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:AULZ3AZC67UNKGCRSKUH5RF5GP","target":"record","payload":{"canonical_record":{"source":{"id":"2602.15338","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-17T03:58:55Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"892ca8ca3244d3ec44dd8ab6b7cd228fef9aa52838a4c082d99c5a212b62e849","abstract_canon_sha256":"e36bae88821557c6b522dc193d2b16e86aa2ee353bdf41e23bf18704e5721a6e"},"schema_version":"1.0"},"canonical_sha256":"05179d8322f7e8d5185192a87ec4bd33d537203a2c3e3097dad012a3b01180fb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:03:18.154712Z","signature_b64":"votIQ+T5hvKHDBsI/lcvdngkeKB8AWKSpFUnkeS+Jw8dmvM2gIdTJX/mRygsKjmJxLglnzqfrG50yxry/SzZAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"05179d8322f7e8d5185192a87ec4bd33d537203a2c3e3097dad012a3b01180fb","last_reissued_at":"2026-05-22T01:03:18.153860Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:03:18.153860Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.15338","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":"2026-05-22T01:03:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hYcE2iE4gyOXAHaO4xQ4GKiniJjJ/XcFMlBZXeTkNyc5wzbqVsSJA036NsBKDXn3RsmuPS+QluZiYeFRIp/cAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T23:02:42.570172Z"},"content_sha256":"2ddc532f20f64bac76aeb17ff278051cef9cf67810eb37df6feb183534a65039","schema_version":"1.0","event_id":"sha256:2ddc532f20f64bac76aeb17ff278051cef9cf67810eb37df6feb183534a65039"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:AULZ3AZC67UNKGCRSKUH5RF5GP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Discovering Implicit Large Language Model Alignment Objectives","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Carlos Guestrin, Edward Chen, Sanmi Koyejo","submitted_at":"2026-02-17T03:58:55Z","abstract_excerpt":"Large language model (LLM) alignment relies on complex reward signals that often obscure the specific behaviors being incentivized, creating critical risks of misalignment and reward hacking. Existing interpretation methods typically rely on pre-defined rubrics, risking the omission of \"unknown unknowns\", or fail to identify objectives that comprehensively cover and are causal to the model behavior. To address these limitations, we introduce Obj-Disco, a framework that automatically decomposes an alignment reward signal into a sparse, weighted combination of human-interpretable natural languag"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.15338","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/2602.15338/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":"2026-05-22T01:03:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GRbdvZW0l0P+ahMzMx9HsAtpsW0Onojer/N9ejgh2O/oErg9nH2icCFkXit+G4KwxzcfeCvn3B5sPYNzaW3cBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T23:02:42.570914Z"},"content_sha256":"29442ef69e10910b8f4c751e96462521b75f5d92c63cff110f77355856c7a882","schema_version":"1.0","event_id":"sha256:29442ef69e10910b8f4c751e96462521b75f5d92c63cff110f77355856c7a882"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AULZ3AZC67UNKGCRSKUH5RF5GP/bundle.json","state_url":"https://pith.science/pith/AULZ3AZC67UNKGCRSKUH5RF5GP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AULZ3AZC67UNKGCRSKUH5RF5GP/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":"2026-05-28T23:02:42Z","links":{"resolver":"https://pith.science/pith/AULZ3AZC67UNKGCRSKUH5RF5GP","bundle":"https://pith.science/pith/AULZ3AZC67UNKGCRSKUH5RF5GP/bundle.json","state":"https://pith.science/pith/AULZ3AZC67UNKGCRSKUH5RF5GP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AULZ3AZC67UNKGCRSKUH5RF5GP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:AULZ3AZC67UNKGCRSKUH5RF5GP","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":"e36bae88821557c6b522dc193d2b16e86aa2ee353bdf41e23bf18704e5721a6e","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-17T03:58:55Z","title_canon_sha256":"892ca8ca3244d3ec44dd8ab6b7cd228fef9aa52838a4c082d99c5a212b62e849"},"schema_version":"1.0","source":{"id":"2602.15338","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.15338","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"arxiv_version","alias_value":"2602.15338v2","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.15338","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"pith_short_12","alias_value":"AULZ3AZC67UN","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"pith_short_16","alias_value":"AULZ3AZC67UNKGCR","created_at":"2026-05-22T01:03:18Z"},{"alias_kind":"pith_short_8","alias_value":"AULZ3AZC","created_at":"2026-05-22T01:03:18Z"}],"graph_snapshots":[{"event_id":"sha256:29442ef69e10910b8f4c751e96462521b75f5d92c63cff110f77355856c7a882","target":"graph","created_at":"2026-05-22T01:03:18Z","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/2602.15338/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language model (LLM) alignment relies on complex reward signals that often obscure the specific behaviors being incentivized, creating critical risks of misalignment and reward hacking. Existing interpretation methods typically rely on pre-defined rubrics, risking the omission of \"unknown unknowns\", or fail to identify objectives that comprehensively cover and are causal to the model behavior. To address these limitations, we introduce Obj-Disco, a framework that automatically decomposes an alignment reward signal into a sparse, weighted combination of human-interpretable natural languag","authors_text":"Carlos Guestrin, Edward Chen, Sanmi Koyejo","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-17T03:58:55Z","title":"Discovering Implicit Large Language Model Alignment Objectives"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.15338","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":"sha256:2ddc532f20f64bac76aeb17ff278051cef9cf67810eb37df6feb183534a65039","target":"record","created_at":"2026-05-22T01:03:18Z","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":"e36bae88821557c6b522dc193d2b16e86aa2ee353bdf41e23bf18704e5721a6e","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-17T03:58:55Z","title_canon_sha256":"892ca8ca3244d3ec44dd8ab6b7cd228fef9aa52838a4c082d99c5a212b62e849"},"schema_version":"1.0","source":{"id":"2602.15338","kind":"arxiv","version":2}},"canonical_sha256":"05179d8322f7e8d5185192a87ec4bd33d537203a2c3e3097dad012a3b01180fb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"05179d8322f7e8d5185192a87ec4bd33d537203a2c3e3097dad012a3b01180fb","first_computed_at":"2026-05-22T01:03:18.153860Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-22T01:03:18.153860Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"votIQ+T5hvKHDBsI/lcvdngkeKB8AWKSpFUnkeS+Jw8dmvM2gIdTJX/mRygsKjmJxLglnzqfrG50yxry/SzZAg==","signature_status":"signed_v1","signed_at":"2026-05-22T01:03:18.154712Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.15338","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2ddc532f20f64bac76aeb17ff278051cef9cf67810eb37df6feb183534a65039","sha256:29442ef69e10910b8f4c751e96462521b75f5d92c63cff110f77355856c7a882"],"state_sha256":"997b9ba6eb7c952102f4d603a01c73fa19241e608c161a6756c04815f6644f60"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OXhssI15+0LEsDTMMSvBCxcgN1lDNgEq/Cjv55sihD8/02BuW+2S4o3tVjLtSATDmhefx7RoyMYNZL/qvD6RDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T23:02:42.574831Z","bundle_sha256":"bb681df1198df61e7719f06d1bf5e09d7999e672ec8b3e626ef9ebf041fa0dd3"}}