{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:T4MAKGGRUSXVU4JYCLCFQU7DLO","short_pith_number":"pith:T4MAKGGR","canonical_record":{"source":{"id":"2606.29758","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-29T04:04:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"87f889aa9c35e2edfad5db528edd5a5c4ba5c9e7d3ac0e9c2b956d8ed26a0982","abstract_canon_sha256":"28ce8172cb9866a5c2f504e1c5ed3c42d5cc2b6558a771146d51a106a4e24ac0"},"schema_version":"1.0"},"canonical_sha256":"9f180518d1a4af5a713812c45853e35b8e00a0b15d5868b0d6b3043bb684f01e","source":{"kind":"arxiv","id":"2606.29758","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.29758","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"arxiv_version","alias_value":"2606.29758v1","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29758","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"pith_short_12","alias_value":"T4MAKGGRUSXV","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"pith_short_16","alias_value":"T4MAKGGRUSXVU4JY","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"pith_short_8","alias_value":"T4MAKGGR","created_at":"2026-06-30T02:17:33Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:T4MAKGGRUSXVU4JYCLCFQU7DLO","target":"record","payload":{"canonical_record":{"source":{"id":"2606.29758","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-29T04:04:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"87f889aa9c35e2edfad5db528edd5a5c4ba5c9e7d3ac0e9c2b956d8ed26a0982","abstract_canon_sha256":"28ce8172cb9866a5c2f504e1c5ed3c42d5cc2b6558a771146d51a106a4e24ac0"},"schema_version":"1.0"},"canonical_sha256":"9f180518d1a4af5a713812c45853e35b8e00a0b15d5868b0d6b3043bb684f01e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T02:17:33.911019Z","signature_b64":"+2QH7FADaCPGrbxPB0ypbOeLJe4ft8/wtjXmIYZMD/gHmWdnXRLniUH6ejbIB8dm9Ct1hZUkIR/BXQ6rO68mBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9f180518d1a4af5a713812c45853e35b8e00a0b15d5868b0d6b3043bb684f01e","last_reissued_at":"2026-06-30T02:17:33.910546Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T02:17:33.910546Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.29758","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":"2026-06-30T02:17:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EoQCi+XBOugcJ9nYbECBpcb3uAQznOm5IMd72TYwyMppvoeB9JKbfHT4UUw8VV6cV3ZythgNP8ok+C168JTfDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T19:01:16.626767Z"},"content_sha256":"b704e4bd55fe280641897335bf2b5aa53884c710aaf827e56a3fdf527797ef0e","schema_version":"1.0","event_id":"sha256:b704e4bd55fe280641897335bf2b5aa53884c710aaf827e56a3fdf527797ef0e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:T4MAKGGRUSXVU4JYCLCFQU7DLO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"PS-PPO: Prefix-Sampling PPO for Critic-Free RLHF","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Doo Hwan Hwang, Kee-Eung Kim","submitted_at":"2026-06-29T04:04:32Z","abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) for Large Language Models increasingly relies on critic-free methods as a practical alternative to actor--critic training. Despite their simplicity, existing critic-free approaches propagate a trajectory-level learning signal uniformly across all tokens in a trajectory. This requires full-trajectory policy updates for every rollout, leading to substantial optimization cost for long reasoning traces, even though intermediate prefixes often contain enough information to largely determine the final outcome. We propose Prefix-Sampling Proximal Poli"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29758","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/2606.29758/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":"2026-06-30T02:17:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"I59jmWuuXZqj13fW/jHSTuLRtN3NrtsjAani7hxPNzmDHbNhG+GctEZmwLo+Hylw0qkf5nMvU2GlMvZcE2/KCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T19:01:16.627133Z"},"content_sha256":"7cf5686356ddb51696adeff2e16d1c605c33f50ccfc6ff214e6309f3e774df7a","schema_version":"1.0","event_id":"sha256:7cf5686356ddb51696adeff2e16d1c605c33f50ccfc6ff214e6309f3e774df7a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO/bundle.json","state_url":"https://pith.science/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO/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":"2026-06-30T19:01:16Z","links":{"resolver":"https://pith.science/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO","bundle":"https://pith.science/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO/bundle.json","state":"https://pith.science/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/T4MAKGGRUSXVU4JYCLCFQU7DLO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:T4MAKGGRUSXVU4JYCLCFQU7DLO","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":"28ce8172cb9866a5c2f504e1c5ed3c42d5cc2b6558a771146d51a106a4e24ac0","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-29T04:04:32Z","title_canon_sha256":"87f889aa9c35e2edfad5db528edd5a5c4ba5c9e7d3ac0e9c2b956d8ed26a0982"},"schema_version":"1.0","source":{"id":"2606.29758","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.29758","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"arxiv_version","alias_value":"2606.29758v1","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29758","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"pith_short_12","alias_value":"T4MAKGGRUSXV","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"pith_short_16","alias_value":"T4MAKGGRUSXVU4JY","created_at":"2026-06-30T02:17:33Z"},{"alias_kind":"pith_short_8","alias_value":"T4MAKGGR","created_at":"2026-06-30T02:17:33Z"}],"graph_snapshots":[{"event_id":"sha256:7cf5686356ddb51696adeff2e16d1c605c33f50ccfc6ff214e6309f3e774df7a","target":"graph","created_at":"2026-06-30T02:17:33Z","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/2606.29758/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) for Large Language Models increasingly relies on critic-free methods as a practical alternative to actor--critic training. Despite their simplicity, existing critic-free approaches propagate a trajectory-level learning signal uniformly across all tokens in a trajectory. This requires full-trajectory policy updates for every rollout, leading to substantial optimization cost for long reasoning traces, even though intermediate prefixes often contain enough information to largely determine the final outcome. We propose Prefix-Sampling Proximal Poli","authors_text":"Doo Hwan Hwang, Kee-Eung Kim","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-29T04:04:32Z","title":"PS-PPO: Prefix-Sampling PPO for Critic-Free RLHF"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29758","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:b704e4bd55fe280641897335bf2b5aa53884c710aaf827e56a3fdf527797ef0e","target":"record","created_at":"2026-06-30T02:17:33Z","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":"28ce8172cb9866a5c2f504e1c5ed3c42d5cc2b6558a771146d51a106a4e24ac0","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-29T04:04:32Z","title_canon_sha256":"87f889aa9c35e2edfad5db528edd5a5c4ba5c9e7d3ac0e9c2b956d8ed26a0982"},"schema_version":"1.0","source":{"id":"2606.29758","kind":"arxiv","version":1}},"canonical_sha256":"9f180518d1a4af5a713812c45853e35b8e00a0b15d5868b0d6b3043bb684f01e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9f180518d1a4af5a713812c45853e35b8e00a0b15d5868b0d6b3043bb684f01e","first_computed_at":"2026-06-30T02:17:33.910546Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-30T02:17:33.910546Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+2QH7FADaCPGrbxPB0ypbOeLJe4ft8/wtjXmIYZMD/gHmWdnXRLniUH6ejbIB8dm9Ct1hZUkIR/BXQ6rO68mBw==","signature_status":"signed_v1","signed_at":"2026-06-30T02:17:33.911019Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.29758","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b704e4bd55fe280641897335bf2b5aa53884c710aaf827e56a3fdf527797ef0e","sha256:7cf5686356ddb51696adeff2e16d1c605c33f50ccfc6ff214e6309f3e774df7a"],"state_sha256":"83e5486748563600300dead8f550bc07a960307aa5d0c2d12f17ea0e7bf38531"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1mRHI9MfO9ktqj4fykbsLx9rxY9QX6TKcoTNPj+h8UbQfQ8rbjTAjSawmoylwtYlh8Mrpc9ctcWbiatQ7i8XDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-30T19:01:16.629090Z","bundle_sha256":"fba5c5aa13bff016c20e4d48155a7635d4815abb5d57c30372b4968cae30e8aa"}}