{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:YTVW6VQVBTME74PSRGM52KSK76","short_pith_number":"pith:YTVW6VQV","canonical_record":{"source":{"id":"2412.10529","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-13T19:47:38Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"4d31c95b7dee729415a2cabddde4c19e257289ba25daf805d86f67e212c1b349","abstract_canon_sha256":"6a55f460f574ec0ee0e50e58a75ae25a16a60c72636112e1a0d2532fe0f498da"},"schema_version":"1.0"},"canonical_sha256":"c4eb6f56150cd84ff1f28999dd2a4affb297d29a68c6c1104d51f5f6c7824aec","source":{"kind":"arxiv","id":"2412.10529","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.10529","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"arxiv_version","alias_value":"2412.10529v1","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.10529","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"pith_short_12","alias_value":"YTVW6VQVBTME","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"pith_short_16","alias_value":"YTVW6VQVBTME74PS","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"pith_short_8","alias_value":"YTVW6VQV","created_at":"2026-07-05T09:49:17Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:YTVW6VQVBTME74PSRGM52KSK76","target":"record","payload":{"canonical_record":{"source":{"id":"2412.10529","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-13T19:47:38Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"4d31c95b7dee729415a2cabddde4c19e257289ba25daf805d86f67e212c1b349","abstract_canon_sha256":"6a55f460f574ec0ee0e50e58a75ae25a16a60c72636112e1a0d2532fe0f498da"},"schema_version":"1.0"},"canonical_sha256":"c4eb6f56150cd84ff1f28999dd2a4affb297d29a68c6c1104d51f5f6c7824aec","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:49:17.246699Z","signature_b64":"9ZIZHiAUlmN7BDazqFSQJAZGNbmUg5WVv9I5kukVHBB1accSEYlO/ZNxxn1tTcmxamVzj+T+BohA+5xzr3QlDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c4eb6f56150cd84ff1f28999dd2a4affb297d29a68c6c1104d51f5f6c7824aec","last_reissued_at":"2026-07-05T09:49:17.246278Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:49:17.246278Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.10529","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":"2026-07-05T09:49:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IZ9j8R9nWJ1wIbQw5gVd7bOLWYGjr1X3aUxPbwKLY5oDo9xB8okVz7I89L0BiIL/g9fbFGvMW6D1OUoAIpPfDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T03:25:08.062014Z"},"content_sha256":"189fb14628baa243dfefdec4de0e07073a1657126f742f4865e0e272844bb88a","schema_version":"1.0","event_id":"sha256:189fb14628baa243dfefdec4de0e07073a1657126f742f4865e0e272844bb88a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:YTVW6VQVBTME74PSRGM52KSK76","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Solving the Inverse Alignment Problem for Efficient RLHF","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Aishwarya Sahoo, Shambhavi Krishna","submitted_at":"2024-12-13T19:47:38Z","abstract_excerpt":"Collecting high-quality preference datasets for reinforcement learning from human feedback (RLHF) is resource-intensive and challenging. As a result, researchers often train reward models on extensive offline datasets which aggregate diverse generation sources and scoring/alignment policies. We hypothesize that this aggregation has an averaging effect on reward model scores, which limits signal and impairs the alignment process. Inspired by the field of inverse RL, we define the 'inverse alignment problem' in language model training, where our objective is to optimize the critic's reward for a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.10529","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/2412.10529/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":"2026-07-05T09:49:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Oy6bM3MCRSrgU6fDfJZHNlxJgMQpY0ki+vWReFXcLrnMdE8dX09bohKeafcAJxI51F2BcB1YdMZ0H5haF3McDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T03:25:08.062386Z"},"content_sha256":"7853b465971c8c96e7900818bcdcc9971ce34453c59425e17ff5506004ee2ace","schema_version":"1.0","event_id":"sha256:7853b465971c8c96e7900818bcdcc9971ce34453c59425e17ff5506004ee2ace"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YTVW6VQVBTME74PSRGM52KSK76/bundle.json","state_url":"https://pith.science/pith/YTVW6VQVBTME74PSRGM52KSK76/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YTVW6VQVBTME74PSRGM52KSK76/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":"2026-08-19T03:25:08Z","links":{"resolver":"https://pith.science/pith/YTVW6VQVBTME74PSRGM52KSK76","bundle":"https://pith.science/pith/YTVW6VQVBTME74PSRGM52KSK76/bundle.json","state":"https://pith.science/pith/YTVW6VQVBTME74PSRGM52KSK76/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YTVW6VQVBTME74PSRGM52KSK76/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YTVW6VQVBTME74PSRGM52KSK76","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":"6a55f460f574ec0ee0e50e58a75ae25a16a60c72636112e1a0d2532fe0f498da","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-13T19:47:38Z","title_canon_sha256":"4d31c95b7dee729415a2cabddde4c19e257289ba25daf805d86f67e212c1b349"},"schema_version":"1.0","source":{"id":"2412.10529","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.10529","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"arxiv_version","alias_value":"2412.10529v1","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.10529","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"pith_short_12","alias_value":"YTVW6VQVBTME","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"pith_short_16","alias_value":"YTVW6VQVBTME74PS","created_at":"2026-07-05T09:49:17Z"},{"alias_kind":"pith_short_8","alias_value":"YTVW6VQV","created_at":"2026-07-05T09:49:17Z"}],"graph_snapshots":[{"event_id":"sha256:7853b465971c8c96e7900818bcdcc9971ce34453c59425e17ff5506004ee2ace","target":"graph","created_at":"2026-07-05T09:49:17Z","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/2412.10529/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Collecting high-quality preference datasets for reinforcement learning from human feedback (RLHF) is resource-intensive and challenging. As a result, researchers often train reward models on extensive offline datasets which aggregate diverse generation sources and scoring/alignment policies. We hypothesize that this aggregation has an averaging effect on reward model scores, which limits signal and impairs the alignment process. Inspired by the field of inverse RL, we define the 'inverse alignment problem' in language model training, where our objective is to optimize the critic's reward for a","authors_text":"Aishwarya Sahoo, Shambhavi Krishna","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-13T19:47:38Z","title":"Solving the Inverse Alignment Problem for Efficient RLHF"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.10529","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":"sha256:189fb14628baa243dfefdec4de0e07073a1657126f742f4865e0e272844bb88a","target":"record","created_at":"2026-07-05T09:49:17Z","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":"6a55f460f574ec0ee0e50e58a75ae25a16a60c72636112e1a0d2532fe0f498da","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-13T19:47:38Z","title_canon_sha256":"4d31c95b7dee729415a2cabddde4c19e257289ba25daf805d86f67e212c1b349"},"schema_version":"1.0","source":{"id":"2412.10529","kind":"arxiv","version":1}},"canonical_sha256":"c4eb6f56150cd84ff1f28999dd2a4affb297d29a68c6c1104d51f5f6c7824aec","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c4eb6f56150cd84ff1f28999dd2a4affb297d29a68c6c1104d51f5f6c7824aec","first_computed_at":"2026-07-05T09:49:17.246278Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:49:17.246278Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"9ZIZHiAUlmN7BDazqFSQJAZGNbmUg5WVv9I5kukVHBB1accSEYlO/ZNxxn1tTcmxamVzj+T+BohA+5xzr3QlDw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:49:17.246699Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.10529","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:189fb14628baa243dfefdec4de0e07073a1657126f742f4865e0e272844bb88a","sha256:7853b465971c8c96e7900818bcdcc9971ce34453c59425e17ff5506004ee2ace"],"state_sha256":"91162bec4d9bb6c0bd4e2eac94b464d8eed455841a35d4ec33d574566b7c2a49"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yplXp1+q99ofoqvdepuFf7hhWNrn+EGJ8DXI5lR0NQ46PIktL1uHl9xdxLzbNxXhPZ1Aq0jskxpWyLf5SaPSCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T03:25:08.065799Z","bundle_sha256":"f884b8db52c1c5952d2dfac17b10e32bee7623fc957a18190ac5604c8e65c5cc"}}