{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:75NYY2KQ5I6LXYHTMCJCYJTTVP","short_pith_number":"pith:75NYY2KQ","schema_version":"1.0","canonical_sha256":"ff5b8c6950ea3cbbe0f360922c2673abece7c42cb03a6b0779bcb40223fc712f","source":{"kind":"arxiv","id":"2606.28707","version":1},"attestation_state":"computed","paper":{"title":"BV-Blend: Uncertainty-Weighted Historical Baselines for Stable Critic-Free RL with Verifiable Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Yi Chang, Yuan Wu, Yupeng Chang","submitted_at":"2026-06-27T03:25:53Z","abstract_excerpt":"Critic-free reinforcement learning with verifiable rewards (RLVR), exemplified by Group Relative Policy Optimization (GRPO), avoids training a value function (critic) and reduces memory and compute overhead relative to critic-based PPO pipelines for aligning large language models. However, GRPO-style advantage estimation depends on prompt-local (within-prompt-group) reward statistics and can be unstable. In particular, when all rollouts in a prompt group receive identical rewards, the within-group reward variance becomes zero, and group normalization yields zero advantages for that group, impe"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.28707","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-27T03:25:53Z","cross_cats_sorted":[],"title_canon_sha256":"aa6cac7b5d49e80df38dc99b21399f9d1e5058872a1605a6f037f9353d71b247","abstract_canon_sha256":"953cb4710834ab4e317b577b2222834af4924bbd363a18e0b23c686e0282e65a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:16:48.742559Z","signature_b64":"4DT+ydwxqjopnPY1sH3uFAtcyU9++UXYuqwnzdr4p90HdnyYnnD5ignXAfHA2Kl+o4UBsklNL2qWP3f5BKCzDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ff5b8c6950ea3cbbe0f360922c2673abece7c42cb03a6b0779bcb40223fc712f","last_reissued_at":"2026-06-30T01:16:48.742046Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:16:48.742046Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BV-Blend: Uncertainty-Weighted Historical Baselines for Stable Critic-Free RL with Verifiable Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Yi Chang, Yuan Wu, Yupeng Chang","submitted_at":"2026-06-27T03:25:53Z","abstract_excerpt":"Critic-free reinforcement learning with verifiable rewards (RLVR), exemplified by Group Relative Policy Optimization (GRPO), avoids training a value function (critic) and reduces memory and compute overhead relative to critic-based PPO pipelines for aligning large language models. However, GRPO-style advantage estimation depends on prompt-local (within-prompt-group) reward statistics and can be unstable. In particular, when all rollouts in a prompt group receive identical rewards, the within-group reward variance becomes zero, and group normalization yields zero advantages for that group, impe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.28707","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/2606.28707/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.28707","created_at":"2026-06-30T01:16:48.742121+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.28707v1","created_at":"2026-06-30T01:16:48.742121+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.28707","created_at":"2026-06-30T01:16:48.742121+00:00"},{"alias_kind":"pith_short_12","alias_value":"75NYY2KQ5I6L","created_at":"2026-06-30T01:16:48.742121+00:00"},{"alias_kind":"pith_short_16","alias_value":"75NYY2KQ5I6LXYHT","created_at":"2026-06-30T01:16:48.742121+00:00"},{"alias_kind":"pith_short_8","alias_value":"75NYY2KQ","created_at":"2026-06-30T01:16:48.742121+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP","json":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP.json","graph_json":"https://pith.science/api/pith-number/75NYY2KQ5I6LXYHTMCJCYJTTVP/graph.json","events_json":"https://pith.science/api/pith-number/75NYY2KQ5I6LXYHTMCJCYJTTVP/events.json","paper":"https://pith.science/paper/75NYY2KQ"},"agent_actions":{"view_html":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP","download_json":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP.json","view_paper":"https://pith.science/paper/75NYY2KQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.28707&json=true","fetch_graph":"https://pith.science/api/pith-number/75NYY2KQ5I6LXYHTMCJCYJTTVP/graph.json","fetch_events":"https://pith.science/api/pith-number/75NYY2KQ5I6LXYHTMCJCYJTTVP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP/action/storage_attestation","attest_author":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP/action/author_attestation","sign_citation":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP/action/citation_signature","submit_replication":"https://pith.science/pith/75NYY2KQ5I6LXYHTMCJCYJTTVP/action/replication_record"}},"created_at":"2026-06-30T01:16:48.742121+00:00","updated_at":"2026-06-30T01:16:48.742121+00:00"}