{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:DABNTJGSNXDGJJTTQ5BX6XIWJQ","short_pith_number":"pith:DABNTJGS","canonical_record":{"source":{"id":"2506.12446","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-14T10:58:38Z","cross_cats_sorted":[],"title_canon_sha256":"1c8587baf7f46ccb534a302a9ccae8b441912891ece43cfed537d80de95e0f04","abstract_canon_sha256":"93246cc193af6b473a99cec81dd11d0f6d1855f70a680811f47999b73871316c"},"schema_version":"1.0"},"canonical_sha256":"1802d9a4d26dc664a67387437f5d164c1648a7231e5a2d101e687eedc069513a","source":{"kind":"arxiv","id":"2506.12446","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.12446","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"arxiv_version","alias_value":"2506.12446v2","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.12446","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"pith_short_12","alias_value":"DABNTJGSNXDG","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"pith_short_16","alias_value":"DABNTJGSNXDGJJTT","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"pith_short_8","alias_value":"DABNTJGS","created_at":"2026-07-05T11:28:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:DABNTJGSNXDGJJTTQ5BX6XIWJQ","target":"record","payload":{"canonical_record":{"source":{"id":"2506.12446","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-14T10:58:38Z","cross_cats_sorted":[],"title_canon_sha256":"1c8587baf7f46ccb534a302a9ccae8b441912891ece43cfed537d80de95e0f04","abstract_canon_sha256":"93246cc193af6b473a99cec81dd11d0f6d1855f70a680811f47999b73871316c"},"schema_version":"1.0"},"canonical_sha256":"1802d9a4d26dc664a67387437f5d164c1648a7231e5a2d101e687eedc069513a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:28:37.457598Z","signature_b64":"qYGLE2P4E2exIRNLU8ufL45hCXwdScc3Z9edu50zPyJZAazwqDfxdwVX5A4kCvEwjtSdxNiQx4QDxP4NMMQ/Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1802d9a4d26dc664a67387437f5d164c1648a7231e5a2d101e687eedc069513a","last_reissued_at":"2026-07-05T11:28:37.457089Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:28:37.457089Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.12446","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":"2026-07-05T11:28:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IcTp+GnQLxXbu9j+I8JIAUThTq3n6JWaD31TI7rJEKoRbx0baNTIOmLNObmVTEDua82107emvHtDcN8/iD6CBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T08:03:43.499835Z"},"content_sha256":"9d4bdf7bf9117498792145aa4b17700f9c4f6b376f125f6a1074907e4dcc1292","schema_version":"1.0","event_id":"sha256:9d4bdf7bf9117498792145aa4b17700f9c4f6b376f125f6a1074907e4dcc1292"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:DABNTJGSNXDGJJTTQ5BX6XIWJQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bingbing Xu, Bin Xie, Huawei Shen, Shengmao Zhu, Yige Yuan","submitted_at":"2025-06-14T10:58:38Z","abstract_excerpt":"Inference-time alignment methods have gained significant attention for their efficiency and effectiveness in aligning large language models (LLMs) with human preferences. However, existing dominant approaches using reward-guided search (RGS) primarily rely on outcome reward models (ORMs), which suffer from a critical granularity mismatch: ORMs are designed to provide outcome rewards for complete responses, while RGS methods rely on process rewards to guide the policy, leading to inconsistent scoring and suboptimal alignment. To address this challenge, we introduce process reward models (PRMs) "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.12446","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/2506.12446/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":"2026-07-05T11:28:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HUO1TpBoFw1FV1/gsoOpnA0KpizZt+8yTO/kprKs3H7zxExX8VoENywv3pidvv3k1xf5w9fULVAaKVK0qJ38Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T08:03:43.500379Z"},"content_sha256":"a3d417e122f35680cc520b88818eaf1ef52fde68b881b092e69ba52672e00186","schema_version":"1.0","event_id":"sha256:a3d417e122f35680cc520b88818eaf1ef52fde68b881b092e69ba52672e00186"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ/bundle.json","state_url":"https://pith.science/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ/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":"2026-08-17T08:03:43Z","links":{"resolver":"https://pith.science/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ","bundle":"https://pith.science/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ/bundle.json","state":"https://pith.science/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DABNTJGSNXDGJJTTQ5BX6XIWJQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DABNTJGSNXDGJJTTQ5BX6XIWJQ","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":"93246cc193af6b473a99cec81dd11d0f6d1855f70a680811f47999b73871316c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-14T10:58:38Z","title_canon_sha256":"1c8587baf7f46ccb534a302a9ccae8b441912891ece43cfed537d80de95e0f04"},"schema_version":"1.0","source":{"id":"2506.12446","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.12446","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"arxiv_version","alias_value":"2506.12446v2","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.12446","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"pith_short_12","alias_value":"DABNTJGSNXDG","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"pith_short_16","alias_value":"DABNTJGSNXDGJJTT","created_at":"2026-07-05T11:28:37Z"},{"alias_kind":"pith_short_8","alias_value":"DABNTJGS","created_at":"2026-07-05T11:28:37Z"}],"graph_snapshots":[{"event_id":"sha256:a3d417e122f35680cc520b88818eaf1ef52fde68b881b092e69ba52672e00186","target":"graph","created_at":"2026-07-05T11:28:37Z","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/2506.12446/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Inference-time alignment methods have gained significant attention for their efficiency and effectiveness in aligning large language models (LLMs) with human preferences. However, existing dominant approaches using reward-guided search (RGS) primarily rely on outcome reward models (ORMs), which suffer from a critical granularity mismatch: ORMs are designed to provide outcome rewards for complete responses, while RGS methods rely on process rewards to guide the policy, leading to inconsistent scoring and suboptimal alignment. To address this challenge, we introduce process reward models (PRMs) ","authors_text":"Bingbing Xu, Bin Xie, Huawei Shen, Shengmao Zhu, Yige Yuan","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-14T10:58:38Z","title":"From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.12446","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":"sha256:9d4bdf7bf9117498792145aa4b17700f9c4f6b376f125f6a1074907e4dcc1292","target":"record","created_at":"2026-07-05T11:28:37Z","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":"93246cc193af6b473a99cec81dd11d0f6d1855f70a680811f47999b73871316c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-14T10:58:38Z","title_canon_sha256":"1c8587baf7f46ccb534a302a9ccae8b441912891ece43cfed537d80de95e0f04"},"schema_version":"1.0","source":{"id":"2506.12446","kind":"arxiv","version":2}},"canonical_sha256":"1802d9a4d26dc664a67387437f5d164c1648a7231e5a2d101e687eedc069513a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1802d9a4d26dc664a67387437f5d164c1648a7231e5a2d101e687eedc069513a","first_computed_at":"2026-07-05T11:28:37.457089Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:28:37.457089Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qYGLE2P4E2exIRNLU8ufL45hCXwdScc3Z9edu50zPyJZAazwqDfxdwVX5A4kCvEwjtSdxNiQx4QDxP4NMMQ/Cg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:28:37.457598Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.12446","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9d4bdf7bf9117498792145aa4b17700f9c4f6b376f125f6a1074907e4dcc1292","sha256:a3d417e122f35680cc520b88818eaf1ef52fde68b881b092e69ba52672e00186"],"state_sha256":"2f8cbc00890bc13841c6a5e2cb706fa48096d6f138e737a4a3497c9d13749192"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"p0Hb/jnQfmILBFz+ZWzbmYvOmnWCLFGS1tEE9qaJ10pOWeqLGKtIT8o1/jfJvgIenFfuYqtuL4inzybt7oGZDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T08:03:43.506210Z","bundle_sha256":"b58b0ff2945c5fac9be4d63e02e23cd33065cab662d3dd0ea5352e82e73d5cd9"}}