{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:25BV3Y7HBUEAM5F5LMZWKMI7OE","short_pith_number":"pith:25BV3Y7H","schema_version":"1.0","canonical_sha256":"d7435de3e70d080674bd5b3365311f711ec971562209f3b1bf67b2ff63e90582","source":{"kind":"arxiv","id":"2606.18910","version":1},"attestation_state":"computed","paper":{"title":"REVES: REvision and VErification--Augmented Training for Test-Time Scaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Amr Sharaf, Arijit Biswas, Hongzhou Lin, Mingyi Hong, Mohammad Ghavamzadeh, Ruida Zhou, Xinyan Zhao, Yuanxin Liu, Zhaoran Wang","submitted_at":"2026-06-17T10:37:23Z","abstract_excerpt":"Test-time scaling via sequential revision has emerged as a powerful paradigm for enhancing Large Language Model (LLM) reasoning. However, standard post-training methods primarily optimize single-shot objectives, creating a fundamental misalignment with multi-step inference dynamics. While recent work treats this as multi-turn reinforcement learning (RL), conventional approaches optimize over the multi-step trajectories directly, failing to further exploit the high-quality mistakes in intermediate steps that model can learn from correcting them. We propose a two-stage iterative framework that a"},"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":"2606.18910","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-17T10:37:23Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"b94b726dbf335a9fd466d550e38a69aacb1d0b631f21bb878a9a75e3f3bf0077","abstract_canon_sha256":"3958ef7841f0a1bcc8a403a4ceef910881b72d946523c6ceb4639e996742bd2e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:11:51.642486Z","signature_b64":"XrtCb/pvZ5AGZ6/5R7Uxrux6p6xU0OT/7PaftE6my069pQEUd6ade1m5z8Ps6QxlNFnZ95s7e6zhcHtizautCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7435de3e70d080674bd5b3365311f711ec971562209f3b1bf67b2ff63e90582","last_reissued_at":"2026-06-19T16:11:51.642043Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:11:51.642043Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"REVES: REvision and VErification--Augmented Training for Test-Time Scaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Amr Sharaf, Arijit Biswas, Hongzhou Lin, Mingyi Hong, Mohammad Ghavamzadeh, Ruida Zhou, Xinyan Zhao, Yuanxin Liu, Zhaoran Wang","submitted_at":"2026-06-17T10:37:23Z","abstract_excerpt":"Test-time scaling via sequential revision has emerged as a powerful paradigm for enhancing Large Language Model (LLM) reasoning. However, standard post-training methods primarily optimize single-shot objectives, creating a fundamental misalignment with multi-step inference dynamics. While recent work treats this as multi-turn reinforcement learning (RL), conventional approaches optimize over the multi-step trajectories directly, failing to further exploit the high-quality mistakes in intermediate steps that model can learn from correcting them. We propose a two-stage iterative framework that a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.18910","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/2606.18910/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":"2606.18910","created_at":"2026-06-19T16:11:51.642103+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.18910v1","created_at":"2026-06-19T16:11:51.642103+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.18910","created_at":"2026-06-19T16:11:51.642103+00:00"},{"alias_kind":"pith_short_12","alias_value":"25BV3Y7HBUEA","created_at":"2026-06-19T16:11:51.642103+00:00"},{"alias_kind":"pith_short_16","alias_value":"25BV3Y7HBUEAM5F5","created_at":"2026-06-19T16:11:51.642103+00:00"},{"alias_kind":"pith_short_8","alias_value":"25BV3Y7H","created_at":"2026-06-19T16:11:51.642103+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/25BV3Y7HBUEAM5F5LMZWKMI7OE","json":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE.json","graph_json":"https://pith.science/api/pith-number/25BV3Y7HBUEAM5F5LMZWKMI7OE/graph.json","events_json":"https://pith.science/api/pith-number/25BV3Y7HBUEAM5F5LMZWKMI7OE/events.json","paper":"https://pith.science/paper/25BV3Y7H"},"agent_actions":{"view_html":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE","download_json":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE.json","view_paper":"https://pith.science/paper/25BV3Y7H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.18910&json=true","fetch_graph":"https://pith.science/api/pith-number/25BV3Y7HBUEAM5F5LMZWKMI7OE/graph.json","fetch_events":"https://pith.science/api/pith-number/25BV3Y7HBUEAM5F5LMZWKMI7OE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE/action/storage_attestation","attest_author":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE/action/author_attestation","sign_citation":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE/action/citation_signature","submit_replication":"https://pith.science/pith/25BV3Y7HBUEAM5F5LMZWKMI7OE/action/replication_record"}},"created_at":"2026-06-19T16:11:51.642103+00:00","updated_at":"2026-06-19T16:11:51.642103+00:00"}