{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:OO2D3RD6XHEATH5NY5DU322KDF","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":"67904bdf7f730bb66fe9d28a566e7454555ba6eba27c2c7f6e7a7cc7701082ef","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-09T15:09:27Z","title_canon_sha256":"ec44830ff7b27ba891b808dca66f374e132a4205f905cac9be95422336277a31"},"schema_version":"1.0","source":{"id":"2504.06958","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.06958","created_at":"2026-05-17T23:38:50Z"},{"alias_kind":"arxiv_version","alias_value":"2504.06958v5","created_at":"2026-05-17T23:38:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.06958","created_at":"2026-05-17T23:38:50Z"},{"alias_kind":"pith_short_12","alias_value":"OO2D3RD6XHEA","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"OO2D3RD6XHEATH5N","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"OO2D3RD6","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:78accb7882d9b95063f60cb8ff0472531ab24daf71186867c419d22ae6649fe3","target":"graph","created_at":"2026-05-17T23:38:50Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Through joint RFT on multiple spatio-temporal perception tasks, we developed VideoChat-R1, a powerful Video MLLM. VideoChat-R1 achieves state-of-the-art spatio-temporal perception, demonstrating significant improvements in tasks like temporal grounding (+31.8) and object tracking (+31.2), while also improving general QA benchmarks."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That carefully designed rule-based rewards focused on temporal associations will produce generalizable improvements in video reasoning without post-hoc tuning or hidden data selection that inflates the reported deltas."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Reinforcement fine-tuning with temporal rewards produces VideoChat-R1, a video MLLM showing large gains on spatio-temporal perception benchmarks such as +31.8 temporal grounding and +31.2 object tracking."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Reinforcement fine-tuning with rule-based temporal rewards creates a video model with state-of-the-art spatio-temporal perception."}],"snapshot_sha256":"1e6ba8c30e20bfae80cc583289f8d278c090c847f62641c85a531daaed64360d"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c514041f82825babf008d72b0e9474228497e59a03ee864cd9356814755bc085"},"paper":{"abstract_excerpt":"Reinforcement Learning (RL) benefits Large Language Models (LLMs) for complex reasoning. Inspired by this, we explore integrating spatio-temporal specific rewards into Multimodal Large Language Models (MLLMs) to address the unique challenges of video understanding, such as long-range temporal associations. This paper investigates how rule-based rewards, particularly temporal ones, can improve video reasoning and their generalizability. Our study proposes Reinforcement Fine-Tuning (RFT) as a data-efficient method to enhance video reasoning on specific tasks without sacrificing original capabili","authors_text":"Desen Meng, Limin Wang, Lu Dong, Xiangyu Zeng, Xinhao Li, Yali Wang, Yinan He, Yi Wang, Yu Qiao, Ziang Yan","cross_cats":[],"headline":"Reinforcement fine-tuning with rule-based temporal rewards creates a video model with state-of-the-art spatio-temporal perception.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-09T15:09:27Z","title":"VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning"},"references":{"count":42,"internal_anchors":19,"resolved_work":42,"sample":[{"cited_arxiv_id":"2502.13923","doi":"","is_internal_anchor":true,"ref_index":1,"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"et al.: FlashVTG: Feature layering and adaptive score handling network for video temporal grounding","work_id":"c5f0f767-bf04-40e1-894c-327d95458eb7","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Boosting the generalization and reasoning of vision language models with curriculum reinforcement learning","work_id":"45c3a58c-d62f-4f0b-84d6-98d7c29563e4","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Open- vlthinker: Complex vision-language reasoning via iterative sft-rl cycles","work_id":"de4c64e1-82b1-4f70-8311-a3539e7bf400","year":2025},{"cited_arxiv_id":"2503.21776","doi":"","is_internal_anchor":true,"ref_index":5,"title":"Video-R1: Reinforcing Video Reasoning in MLLMs","work_id":"0ce88332-564c-4361-8e2a-3850eb1ace9c","year":2025}],"snapshot_sha256":"5d9ce39e683f9d81d550955ea6fc3b6eb7b34451bcb19dbec72ab096f15e4a6e"},"source":{"id":"2504.06958","kind":"arxiv","version":5},"verdict":{"created_at":"2026-05-15T20:52:13.629443Z","id":"46600adc-b342-4d2b-93fa-f09c53eabfa1","model_set":{"reader":"grok-4.3"},"one_line_summary":"Reinforcement fine-tuning with temporal rewards produces VideoChat-R1, a video MLLM showing large gains on spatio-temporal perception benchmarks such as +31.8 temporal grounding and +31.2 object tracking.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Reinforcement fine-tuning with rule-based temporal rewards creates a video model with state-of-the-art spatio-temporal perception.","strongest_claim":"Through joint RFT on multiple spatio-temporal perception tasks, we developed VideoChat-R1, a powerful Video MLLM. VideoChat-R1 achieves state-of-the-art spatio-temporal perception, demonstrating significant improvements in tasks like temporal grounding (+31.8) and object tracking (+31.2), while also improving general QA benchmarks.","weakest_assumption":"That carefully designed rule-based rewards focused on temporal associations will produce generalizable improvements in video reasoning without post-hoc tuning or hidden data selection that inflates the reported deltas."}},"verdict_id":"46600adc-b342-4d2b-93fa-f09c53eabfa1"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:2cf037f5b03872e06abdddf9139e30ff807857620ecce08951ff897ae5c9692f","target":"record","created_at":"2026-05-17T23:38:50Z","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":"67904bdf7f730bb66fe9d28a566e7454555ba6eba27c2c7f6e7a7cc7701082ef","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-09T15:09:27Z","title_canon_sha256":"ec44830ff7b27ba891b808dca66f374e132a4205f905cac9be95422336277a31"},"schema_version":"1.0","source":{"id":"2504.06958","kind":"arxiv","version":5}},"canonical_sha256":"73b43dc47eb9c8099fadc7474deb4a195144c4997a545220f336e76bc37dbcc5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"73b43dc47eb9c8099fadc7474deb4a195144c4997a545220f336e76bc37dbcc5","first_computed_at":"2026-05-17T23:38:50.192174Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:50.192174Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"A5+r6tInSXtq85NBbLVUbFHmxKouySz/qahIl5Iu/c/bXKr0SBXLIE28FzdSiyHul5LwBVzGIJguA2eyulxTAw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:50.192613Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.06958","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2cf037f5b03872e06abdddf9139e30ff807857620ecce08951ff897ae5c9692f","sha256:78accb7882d9b95063f60cb8ff0472531ab24daf71186867c419d22ae6649fe3"],"state_sha256":"bf0fe8c841ab0a708bd174b689f68057bb257b054103cfee65a4594836c10a10"}