{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OCYZW2WGJ3TAQCHDRDADYFENAL","short_pith_number":"pith:OCYZW2WG","schema_version":"1.0","canonical_sha256":"70b19b6ac64ee60808e388c03c148d02d884f91dfbe3eb35f5fc7c09d811dc89","source":{"kind":"arxiv","id":"2503.01785","version":1},"attestation_state":"computed","paper":{"title":"Visual-RFT: Visual Reinforcement Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Visual-RFT lets large vision-language models learn visual tasks from perceptual rewards instead of labeled data.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dahua Lin, Haodong Duan, Jiaqi Wang, Xiaoyi Dong, Yuhang Cao, Yuhang Zang, Zeyi Sun, Ziyu Liu","submitted_at":"2025-03-03T18:16:32Z","abstract_excerpt":"Reinforcement Fine-Tuning (RFT) in Large Reasoning Models like OpenAI o1 learns from feedback on its answers, which is especially useful in applications when fine-tuning data is scarce. Recent open-source work like DeepSeek-R1 demonstrates that reinforcement learning with verifiable reward is one key direction in reproducing o1. While the R1-style model has demonstrated success in language models, its application in multi-modal domains remains under-explored. This work introduces Visual Reinforcement Fine-Tuning (Visual-RFT), which further extends the application areas of RFT on visual tasks. "},"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":true,"formal_links_present":false},"canonical_record":{"source":{"id":"2503.01785","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-03T18:16:32Z","cross_cats_sorted":[],"title_canon_sha256":"4feb024d94b70d42e09917ec358fc74a6a2dbfe6a4d7d6621d55fc13747a0e30","abstract_canon_sha256":"96498b0f0f1524900019ecacd3cffafbb3686a8b0da23ad7467db86874b9071b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T04:29:17.081672Z","signature_b64":"H2yBVUM5xSjXbVwdS53zSf+qodiwYkik7DQZqSi0QZg4qwjvNGW5hp5rfLa8GhjwCk0tOnPewbhndcH9mByaBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"70b19b6ac64ee60808e388c03c148d02d884f91dfbe3eb35f5fc7c09d811dc89","last_reissued_at":"2026-05-18T04:29:17.081188Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T04:29:17.081188Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Visual-RFT: Visual Reinforcement Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Visual-RFT lets large vision-language models learn visual tasks from perceptual rewards instead of labeled data.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dahua Lin, Haodong Duan, Jiaqi Wang, Xiaoyi Dong, Yuhang Cao, Yuhang Zang, Zeyi Sun, Ziyu Liu","submitted_at":"2025-03-03T18:16:32Z","abstract_excerpt":"Reinforcement Fine-Tuning (RFT) in Large Reasoning Models like OpenAI o1 learns from feedback on its answers, which is especially useful in applications when fine-tuning data is scarce. Recent open-source work like DeepSeek-R1 demonstrates that reinforcement learning with verifiable reward is one key direction in reproducing o1. While the R1-style model has demonstrated success in language models, its application in multi-modal domains remains under-explored. This work introduces Visual Reinforcement Fine-Tuning (Visual-RFT), which further extends the application areas of RFT on visual tasks. "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Visual-RFT improves accuracy by 24.3% over the baseline in one-shot fine-grained image classification with around 100 samples and exceeds the baseline by 21.9 on COCO's two-shot setting.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the visual perception verifiable reward functions (e.g., IoU) provide sufficiently dense and unbiased signals to guide policy optimization without introducing new failure modes not present in language-only RFT.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Visual-RFT applies reinforcement learning with verifiable perception rewards to improve large vision-language models on fine-grained classification, few-shot detection, and grounding tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Visual-RFT lets large vision-language models learn visual tasks from perceptual rewards instead of labeled data.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"8076afea17fa7125bd56efb0afa36d68826a481619bf10d58d86d523ea1c3055"},"source":{"id":"2503.01785","kind":"arxiv","version":1},"verdict":{"id":"1829aa0f-ed0e-4ff8-8bd7-dcac36f656d3","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T22:11:14.863124Z","strongest_claim":"Visual-RFT improves accuracy by 24.3% over the baseline in one-shot fine-grained image classification with around 100 samples and exceeds the baseline by 21.9 on COCO's two-shot setting.","one_line_summary":"Visual-RFT applies reinforcement learning with verifiable perception rewards to improve large vision-language models on fine-grained classification, few-shot detection, and grounding tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the visual perception verifiable reward functions (e.g., IoU) provide sufficiently dense and unbiased signals to guide policy optimization without introducing new failure modes not present in language-only RFT.","pith_extraction_headline":"Visual-RFT lets large vision-language models learn visual tasks from perceptual rewards instead of labeled data."},"references":{"count":52,"sample":[{"doi":"","year":null,"title":"Lmrl gym: Benchmarks for multi-turn reinforcement learn- ing with language models","work_id":"bad1baad-5c3f-4456-ac96-29c8f5e78bfb","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"InternLM2 Technical Report","work_id":"dfa13e0e-1c3c-4fb6-943d-a19945bacdbe","ref_index":2,"cited_arxiv_id":"2403.17297","is_internal_anchor":true},{"doi":"","year":2023,"title":"Grounding large language models in interactive environments with on- line reinforcement learning","work_id":"3086d7ff-f5e3-4593-9700-3603bad5be12","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","ref_index":4,"cited_arxiv_id":"2501.12948","is_internal_anchor":true},{"doi":"","year":2019,"title":"Lvis: A dataset for large vocabulary instance segmentation","work_id":"d2430c96-329f-4510-aaa0-74f084edb36d","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":52,"snapshot_sha256":"093104e5a7a3d56c18936c23fdb27b5fd0a3cb53533ee3d4794f03e915dcfa41","internal_anchors":19},"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":"2503.01785","created_at":"2026-05-18T04:29:17.081267+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.01785v1","created_at":"2026-05-18T04:29:17.081267+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.01785","created_at":"2026-05-18T04:29:17.081267+00:00"},{"alias_kind":"pith_short_12","alias_value":"OCYZW2WGJ3TA","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"OCYZW2WGJ3TAQCHD","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"OCYZW2WG","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":55,"internal_anchor_count":55,"sample":[{"citing_arxiv_id":"2505.15616","citing_title":"LENS: Multi-level Evaluation of Multimodal Reasoning with Large Language Models","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2505.23678","citing_title":"Grounded Reinforcement Learning for Visual Reasoning","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15864","citing_title":"Are VLMs Seeing or Just Saying? Uncovering the Illusion of Visual Re-examination","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15951","citing_title":"From Failure to Feedback: Group Revision Unlocks Hard Cases in Object-Level Grounding","ref_index":52,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20165","citing_title":"CaMo: Camera Motion Grounded Evaluation and Training for Vision-Language Models","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20164","citing_title":"Not Every Rubric Teaches Equally: Policy-Aware Rubric Rewards for RLVR","ref_index":31,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16903","citing_title":"WOW-Seg: A Word-free Open World Segmentation Model","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2506.06856","citing_title":"Vision-EKIPL: External Knowledge-Infused Policy Learning for Visual Reasoning","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2506.11991","citing_title":"VGR: Visual Grounded Reasoning","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2506.16796","citing_title":"RealSR-R1: Reinforcement Learning for Real-World Image Super-Resolution with Vision-Language Chain-of-Thought","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2503.17352","citing_title":"OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2507.00748","citing_title":"Improving the Reasoning of Multi-Image Grounding in MLLMs via Reinforcement Learning","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2507.00209","citing_title":"SurgiSR4K: A High-Resolution Endoscopic Video Dataset for Robotic-Assisted Minimally Invasive Procedures","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2507.16815","citing_title":"ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2509.02547","citing_title":"The Landscape of Agentic Reinforcement Learning for LLMs: A Survey","ref_index":231,"is_internal_anchor":true},{"citing_arxiv_id":"2509.22746","citing_title":"Mixture-of-Visual-Thoughts: Exploring Context-Adaptive Reasoning Mode Selection for General Visual Reasoning","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2510.10606","citing_title":"ViSurf: Visual Supervised-and-Reinforcement Fine-Tuning for Large Vision-and-Language Models","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2509.07969","citing_title":"Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Search","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2511.19820","citing_title":"CropVLM: Learning to Zoom for Fine-Grained Vision-Language Perception","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2511.20814","citing_title":"SPHINX: A Synthetic Environment for Visual Perception and Reasoning","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2503.13377","citing_title":"Time-R1: Post-Training Large Vision Language Model for Temporal Video Grounding","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2512.03963","citing_title":"TempR1: Improving Temporal Understanding of MLLMs via Temporal-Aware Multi-Task Reinforcement Learning","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2512.10554","citing_title":"Grounding Everything in Tokens for Multimodal Large Language Models","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2512.13671","citing_title":"AgentIAD: Agentic Industrial Anomaly Detection via Adaptive Memory Augmentation","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2512.16918","citing_title":"AdaTooler-V: Adaptive Tool-Use for Images and Videos","ref_index":39,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL","json":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL.json","graph_json":"https://pith.science/api/pith-number/OCYZW2WGJ3TAQCHDRDADYFENAL/graph.json","events_json":"https://pith.science/api/pith-number/OCYZW2WGJ3TAQCHDRDADYFENAL/events.json","paper":"https://pith.science/paper/OCYZW2WG"},"agent_actions":{"view_html":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL","download_json":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL.json","view_paper":"https://pith.science/paper/OCYZW2WG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.01785&json=true","fetch_graph":"https://pith.science/api/pith-number/OCYZW2WGJ3TAQCHDRDADYFENAL/graph.json","fetch_events":"https://pith.science/api/pith-number/OCYZW2WGJ3TAQCHDRDADYFENAL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL/action/storage_attestation","attest_author":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL/action/author_attestation","sign_citation":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL/action/citation_signature","submit_replication":"https://pith.science/pith/OCYZW2WGJ3TAQCHDRDADYFENAL/action/replication_record"}},"created_at":"2026-05-18T04:29:17.081267+00:00","updated_at":"2026-05-18T04:29:17.081267+00:00"}