{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GUNOICWVT5MDYR27JG3T7XQFNI","short_pith_number":"pith:GUNOICWV","schema_version":"1.0","canonical_sha256":"351ae40ad59f583c475f49b73fde056a06a40e3f16aa12475f819b6e7333b569","source":{"kind":"arxiv","id":"2505.19000","version":1},"attestation_state":"computed","paper":{"title":"VerIPO: Cultivating Long Reasoning in Video-LLMs via Verifier-Gudied Iterative Policy Optimization","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Baotian Hu, Longyue Wang, Min Zhang, Wenhan Luo, Xinyu Chen, Yunxin Li, Zhenyu Liu, Zitao Li","submitted_at":"2025-05-25T06:41:28Z","abstract_excerpt":"Applying Reinforcement Learning (RL) to Video Large Language Models (Video-LLMs) shows significant promise for complex video reasoning. However, popular Reinforcement Fine-Tuning (RFT) methods, such as outcome-based Group Relative Policy Optimization (GRPO), are limited by data preparation bottlenecks (e.g., noise or high cost) and exhibit unstable improvements in the quality of long chain-of-thoughts (CoTs) and downstream performance.To address these limitations, we propose VerIPO, a Verifier-guided Iterative Policy Optimization method designed to gradually improve video LLMs' capacity for ge"},"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":"2505.19000","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T06:41:28Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"dcd0820bc0655515050226c7246e2ef6b251304cd008659264ae12b9ffa64aa9","abstract_canon_sha256":"00f31e2d74bfab057d1a89821c1459c9748543ccfa529d22204e50b017f5d2d9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:23.325577Z","signature_b64":"7t2n+Du0yYlaYBfY3s2yw0iZsA/xfacG0qEYZmX31IiTjsJhffVg7pEJRvDnXXHoBRkKhDHXgQ5yueEf1r0RBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"351ae40ad59f583c475f49b73fde056a06a40e3f16aa12475f819b6e7333b569","last_reissued_at":"2026-07-05T11:09:23.325044Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:23.325044Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VerIPO: Cultivating Long Reasoning in Video-LLMs via Verifier-Gudied Iterative Policy Optimization","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Baotian Hu, Longyue Wang, Min Zhang, Wenhan Luo, Xinyu Chen, Yunxin Li, Zhenyu Liu, Zitao Li","submitted_at":"2025-05-25T06:41:28Z","abstract_excerpt":"Applying Reinforcement Learning (RL) to Video Large Language Models (Video-LLMs) shows significant promise for complex video reasoning. However, popular Reinforcement Fine-Tuning (RFT) methods, such as outcome-based Group Relative Policy Optimization (GRPO), are limited by data preparation bottlenecks (e.g., noise or high cost) and exhibit unstable improvements in the quality of long chain-of-thoughts (CoTs) and downstream performance.To address these limitations, we propose VerIPO, a Verifier-guided Iterative Policy Optimization method designed to gradually improve video LLMs' capacity for ge"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.19000","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/2505.19000/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":"2505.19000","created_at":"2026-07-05T11:09:23.325103+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.19000v1","created_at":"2026-07-05T11:09:23.325103+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.19000","created_at":"2026-07-05T11:09:23.325103+00:00"},{"alias_kind":"pith_short_12","alias_value":"GUNOICWVT5MD","created_at":"2026-07-05T11:09:23.325103+00:00"},{"alias_kind":"pith_short_16","alias_value":"GUNOICWVT5MDYR27","created_at":"2026-07-05T11:09:23.325103+00:00"},{"alias_kind":"pith_short_8","alias_value":"GUNOICWV","created_at":"2026-07-05T11:09:23.325103+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07433","citing_title":"Watch, Remember, Reason: Human-View Video Understanding with MLLMs","ref_index":186,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI","json":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI.json","graph_json":"https://pith.science/api/pith-number/GUNOICWVT5MDYR27JG3T7XQFNI/graph.json","events_json":"https://pith.science/api/pith-number/GUNOICWVT5MDYR27JG3T7XQFNI/events.json","paper":"https://pith.science/paper/GUNOICWV"},"agent_actions":{"view_html":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI","download_json":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI.json","view_paper":"https://pith.science/paper/GUNOICWV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.19000&json=true","fetch_graph":"https://pith.science/api/pith-number/GUNOICWVT5MDYR27JG3T7XQFNI/graph.json","fetch_events":"https://pith.science/api/pith-number/GUNOICWVT5MDYR27JG3T7XQFNI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI/action/storage_attestation","attest_author":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI/action/author_attestation","sign_citation":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI/action/citation_signature","submit_replication":"https://pith.science/pith/GUNOICWVT5MDYR27JG3T7XQFNI/action/replication_record"}},"created_at":"2026-07-05T11:09:23.325103+00:00","updated_at":"2026-07-05T11:09:23.325103+00:00"}