{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DUNYXM5BA4XJRJQEDWYLO2E7N7","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":"ec260685864428546f9219f556523d3969b00abb28dd89bf229c586abfeab891","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-12-04T11:12:13Z","title_canon_sha256":"e7ca40a652a7ac5d695737b9deeff6537d1687485008487091d80ac55bf5c989"},"schema_version":"1.0","source":{"id":"2512.04678","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.04678","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"arxiv_version","alias_value":"2512.04678v2","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.04678","created_at":"2026-05-17T23:38:47Z"},{"alias_kind":"pith_short_12","alias_value":"DUNYXM5BA4XJ","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"DUNYXM5BA4XJRJQE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"DUNYXM5B","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:a62ab4ea916db26cd1900e2ad95209cccecfc1bcc3419ae94fdaec14b8680a5f","target":"graph","created_at":"2026-05-17T23:38:47Z","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":"Reward Forcing achieves state-of-the-art performance on standard benchmarks while enabling high-quality streaming video generation at 23.1 FPS on a single H100 GPU."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The vision-language model used to rate motion dynamics provides an accurate and unbiased signal that genuinely improves the distilled model's motion quality without introducing new artifacts or distribution shifts."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Reward Forcing combines EMA-Sink tokens and Rewarded Distribution Matching Distillation to deliver state-of-the-art streaming video generation at 23.1 FPS without copying initial frames."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"EMA-updated sink tokens and reward-weighted distillation fix copied frames and weak motion in streaming video models."}],"snapshot_sha256":"eb7e58862ca94c53c792d694714e0c207d0eae4f4054ad71aea4b5c949807624"},"formal_canon":{"evidence_count":3,"snapshot_sha256":"d0084abd67152f2afd7971a315008621fb58aa3a673c22319fde42d1fe2122cc"},"paper":{"abstract_excerpt":"Efficient streaming video generation is critical for simulating interactive and dynamic worlds. Existing methods distill few-step video diffusion models with sliding window attention, using initial frames as sink tokens to maintain attention performance and reduce error accumulation. However, video frames become overly dependent on these static tokens, resulting in copied initial frames and diminished motion dynamics. To address this, we introduce Reward Forcing, a novel framework with two key designs. First, we propose EMA-Sink, which maintains fixed-size tokens initialized from initial frame","authors_text":"Haobo Li, Hao Ouyang, Hengyuan Cao, Jiapeng Zhu, Ka Leong Cheng, Min Zhang, Qiuyu Wang, Xing Zhu, Yanhong Zeng, Yujun Shen, Yunhong Lu, Zhipeng Zhang","cross_cats":[],"headline":"EMA-updated sink tokens and reward-weighted distillation fix copied frames and weak motion in streaming video models.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-12-04T11:12:13Z","title":"Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation"},"references":{"count":99,"internal_anchors":33,"resolved_work":99,"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":"Philip J. Ball, Jakob Bauer, Frank Belletti, Bethanie Brown- field, Ariel Ephrat, Shlomi Fruchter, Agrim Gupta, Kris- tian Holsheimer, Aleksander Holynski, Jiri Hron, Christos Kaplanis, Marjorie Limon","work_id":"6283c28b-9a42-43bf-b29d-45a63104f030","year":2025},{"cited_arxiv_id":"2311.15127","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets","work_id":"4f68eada-27e3-437a-a2fe-6e4ca524d0d3","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Video generation models as world simulators","work_id":"775c18a1-70c0-4557-8352-9608c63bdd24","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Dimension-reduction attack! video generative models are experts on controllable image synthesis.arXiv preprint arXiv:2505.23325, 2025a","work_id":"5ef97af7-d902-425c-b24b-7d796c87372c","year":2025}],"snapshot_sha256":"cdddf6d57b186e7fb851e2d8d941173800e8d20858c4671e17be4f0ff0698345"},"source":{"id":"2512.04678","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-16T18:12:24.693015Z","id":"98362fe4-d383-4e49-8cee-bc56ddd5c37b","model_set":{"reader":"grok-4.3"},"one_line_summary":"Reward Forcing combines EMA-Sink tokens and Rewarded Distribution Matching Distillation to deliver state-of-the-art streaming video generation at 23.1 FPS without copying initial frames.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"EMA-updated sink tokens and reward-weighted distillation fix copied frames and weak motion in streaming video models.","strongest_claim":"Reward Forcing achieves state-of-the-art performance on standard benchmarks while enabling high-quality streaming video generation at 23.1 FPS on a single H100 GPU.","weakest_assumption":"The vision-language model used to rate motion dynamics provides an accurate and unbiased signal that genuinely improves the distilled model's motion quality without introducing new artifacts or distribution shifts."}},"verdict_id":"98362fe4-d383-4e49-8cee-bc56ddd5c37b"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:8e1aa439fcd561f18cf523d74b358988fd35848253fa2f09877efc4f273c71be","target":"record","created_at":"2026-05-17T23:38:47Z","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":"ec260685864428546f9219f556523d3969b00abb28dd89bf229c586abfeab891","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-12-04T11:12:13Z","title_canon_sha256":"e7ca40a652a7ac5d695737b9deeff6537d1687485008487091d80ac55bf5c989"},"schema_version":"1.0","source":{"id":"2512.04678","kind":"arxiv","version":2}},"canonical_sha256":"1d1b8bb3a1072e98a6041db0b7689f6ff1667d284b163bd00c2aa568ba216026","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1d1b8bb3a1072e98a6041db0b7689f6ff1667d284b163bd00c2aa568ba216026","first_computed_at":"2026-05-17T23:38:47.001172Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:47.001172Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Jc24fL7sZs0RG4J2NoIU6yTQ59fe7yhtYoLpP1HN0yNMqfGvVpzkPT5yHOprKXe4gRRwgbD2/4xjVTqfzXXTBg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:47.001748Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.04678","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8e1aa439fcd561f18cf523d74b358988fd35848253fa2f09877efc4f273c71be","sha256:a62ab4ea916db26cd1900e2ad95209cccecfc1bcc3419ae94fdaec14b8680a5f"],"state_sha256":"6f6af4331e0596257667d8fd5467f613d8596f1413487cba7ab49d83314c6b8d"}