{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:5D7V4YGGRZJ55XM4MM3HQ4EPCK","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":"c0368509d49130e999ec1b34eab365d03ecba6561008289a37f4f33a83a34f00","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:09:05Z","title_canon_sha256":"3f141cae6fec5a3aa47b0b1a523eb66e10e06cf6edca392e5a22f411cd1f3dab"},"schema_version":"1.0","source":{"id":"2605.21654","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21654","created_at":"2026-05-22T01:03:26Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21654v1","created_at":"2026-05-22T01:03:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21654","created_at":"2026-05-22T01:03:26Z"},{"alias_kind":"pith_short_12","alias_value":"5D7V4YGGRZJ5","created_at":"2026-05-22T01:03:26Z"},{"alias_kind":"pith_short_16","alias_value":"5D7V4YGGRZJ55XM4","created_at":"2026-05-22T01:03:26Z"},{"alias_kind":"pith_short_8","alias_value":"5D7V4YGG","created_at":"2026-05-22T01:03:26Z"}],"graph_snapshots":[{"event_id":"sha256:11abb422753150da479f1764bf8966b5a51304223f4d3206f83bd21dd24f0ed4","target":"graph","created_at":"2026-05-22T01:03:26Z","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/2605.21654/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning substantially improves pretrained language models, but it remains understudied why critic-free methods such as PPO and GRPO work as well as they do, and when they should provide the largest gains. We develop a value-gradient perspective of critic-free RL for LLM post-training. First, under a differentiable rollout and additive-noise parameterization, we show that the actor update is value-gradient-like in expectation: the backward pass propagates costates whose conditional expectation equals the value gradient. Second, for discrete transformer policies, we show that auto","authors_text":"Arip Asadulaev, Daniil Ognev, Karim Salta, Martin Takac","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:09:05Z","title":"Value-Gradient Hypothesis of RL for LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21654","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:f5640c17c7aab76da0a7fffb7376550c46766cd732b45eee40f9804d888cc3cd","target":"record","created_at":"2026-05-22T01:03:26Z","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":"c0368509d49130e999ec1b34eab365d03ecba6561008289a37f4f33a83a34f00","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:09:05Z","title_canon_sha256":"3f141cae6fec5a3aa47b0b1a523eb66e10e06cf6edca392e5a22f411cd1f3dab"},"schema_version":"1.0","source":{"id":"2605.21654","kind":"arxiv","version":1}},"canonical_sha256":"e8ff5e60c68e53dedd9c633678708f12a09686cefed995893efd820c8ea8ab8c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e8ff5e60c68e53dedd9c633678708f12a09686cefed995893efd820c8ea8ab8c","first_computed_at":"2026-05-22T01:03:26.991607Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-22T01:03:26.991607Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"wrDKSmT83XCcaLK9+Z2THtoBt18IWTvKi5dOkm0nvZKE4359bXo/sMbd5i65rmal+zZMBLi6xRU+x1BJoNAIDw==","signature_status":"signed_v1","signed_at":"2026-05-22T01:03:26.992331Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.21654","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f5640c17c7aab76da0a7fffb7376550c46766cd732b45eee40f9804d888cc3cd","sha256:11abb422753150da479f1764bf8966b5a51304223f4d3206f83bd21dd24f0ed4"],"state_sha256":"63dfbb7324a44fce45e7040402ba12fe83c1c31e63071dffed672b192c683c8d"}