{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VOWDRF6XRAOWX3FIFQE4A4XW5Q","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":"30d65e77cdcf7bba6dd8d9414bea4b24b3463382a50313501e9aca6ca83013df","cross_cats_sorted":["cs.RO"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T09:54:31Z","title_canon_sha256":"e67c7c00cd97c2d5db41021910748ae0e91d5ee597a790f73cfb8c8cd9674275"},"schema_version":"1.0","source":{"id":"2605.13276","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13276","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13276v2","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13276","created_at":"2026-05-18T02:44:49Z"},{"alias_kind":"pith_short_12","alias_value":"VOWDRF6XRAOW","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"VOWDRF6XRAOWX3FI","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"VOWDRF6X","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:53e9d9a7313a72e79995c12af06e32f082265cab9fea6262cc5689cbede5ffb9","target":"graph","created_at":"2026-05-18T02:44:49Z","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":"Experiments on benchmarks like LIBERO show that D-VLA significantly outperforms mainstream RL frameworks in throughput and sampling efficiency for billion-parameter VLA models. In trillion-parameter scalability tests, our framework maintains exceptional stability and linear speedup."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The assumption that physically isolating high-frequency training data from low-frequency weight control via Plane Decoupling, combined with the swimlane pipeline, eliminates interference and achieves linear speedup without new bottlenecks in real distributed hardware."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"D-VLA uses plane decoupling and a swimlane pipeline to deliver higher throughput and linear speedup than prior RL frameworks when training billion- and trillion-parameter VLA models on benchmarks like LIBERO."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"D-VLA decouples simulation and optimization planes to deliver linear speedup for trillion-parameter vision-language-action models in distributed RL."}],"snapshot_sha256":"e35cd56a639a7d8617b0e5592108c81a8c8d16ee6c4cbf728b1ceffb8fed0418"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"521c7f2db1f868d8bf55b06c2e2e5ceb7891412dae076442d39fe0fdf562105f"},"paper":{"abstract_excerpt":"The rapid evolution of Embodied AI has enabled Vision-Language-Action (VLA) models to excel in multimodal perception and task execution. However, applying Reinforcement Learning (RL) to these massive models in large-scale distributed environments faces severe systemic bottlenecks, primarily due to the resource conflict between high-fidelity physical simulation and the intensive VRAM/bandwidth demands of deep learning. This conflict often leaves overall throughput constrained by execution-phase inefficiencies. To address these challenges, we propose D-VLA, a high-concurrency, low-latency distri","authors_text":"Haodong Yue, Haoran Sun, Junwu Xiong, Luqiao Wang, Shuai Di, Wen Huang, Xiaolong Xiang, Yicheng Gong, Yongjian Guo, Yucheng Guo, Zhen Sun, Zhong Guan","cross_cats":["cs.RO"],"headline":"D-VLA decouples simulation and optimization planes to deliver linear speedup for trillion-parameter vision-language-action models in distributed RL.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T09:54:31Z","title":"D-VLA: A High-Concurrency Distributed Asynchronous Reinforcement Learning Framework for Vision-Language-Action Models"},"references":{"count":28,"internal_anchors":13,"resolved_work":28,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Sanketi, Grecia Salazar, Michael S","work_id":"3981437e-1b5e-41d1-8d02-40f4186d18ff","year":2023},{"cited_arxiv_id":"2410.24164","doi":"","is_internal_anchor":true,"ref_index":2,"title":"$\\pi_0$: A Vision-Language-Action Flow Model for General Robot Control","work_id":"f790abdc-a796-482f-a40d-f8ee035ecfc2","year":2024},{"cited_arxiv_id":"2510.03342","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Gemini Robotics 1.5: Pushing the Frontier of Generalist Robots with Advanced Embodied Reasoning, Thinking, and Motion Transfer","work_id":"ffe7b7f4-997b-4957-b253-03cfbebf6f9a","year":2025},{"cited_arxiv_id":"2406.09246","doi":"","is_internal_anchor":true,"ref_index":4,"title":"OpenVLA: An Open-Source Vision-Language-Action Model","work_id":"3e7e65c5-5aed-4fe9-8414-2092bcb31cc7","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Smolvla: A vision-language-action model for affordable and efficient robotics, 2025","work_id":"d3061a03-48d4-4b75-9b97-92f97e21eded","year":2025}],"snapshot_sha256":"67e4e266a72abc87ac8ec898be3d8373fc1ee17157322c019a0e31af20cb1608"},"source":{"id":"2605.13276","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-15T06:00:28.408357Z","id":"add5ac7b-c124-48f1-b9d5-7aa3dc4c2ded","model_set":{"reader":"grok-4.3"},"one_line_summary":"D-VLA uses plane decoupling and a swimlane pipeline to deliver higher throughput and linear speedup than prior RL frameworks when training billion- and trillion-parameter VLA models on benchmarks like LIBERO.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"D-VLA decouples simulation and optimization planes to deliver linear speedup for trillion-parameter vision-language-action models in distributed RL.","strongest_claim":"Experiments on benchmarks like LIBERO show that D-VLA significantly outperforms mainstream RL frameworks in throughput and sampling efficiency for billion-parameter VLA models. In trillion-parameter scalability tests, our framework maintains exceptional stability and linear speedup.","weakest_assumption":"The assumption that physically isolating high-frequency training data from low-frequency weight control via Plane Decoupling, combined with the swimlane pipeline, eliminates interference and achieves linear speedup without new bottlenecks in real distributed hardware."}},"verdict_id":"add5ac7b-c124-48f1-b9d5-7aa3dc4c2ded"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:b32f3be5d9e564c00e212866440ef42a8e49a8b7749465474d267a87d5b1f7ca","target":"record","created_at":"2026-05-18T02:44:49Z","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":"30d65e77cdcf7bba6dd8d9414bea4b24b3463382a50313501e9aca6ca83013df","cross_cats_sorted":["cs.RO"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T09:54:31Z","title_canon_sha256":"e67c7c00cd97c2d5db41021910748ae0e91d5ee597a790f73cfb8c8cd9674275"},"schema_version":"1.0","source":{"id":"2605.13276","kind":"arxiv","version":2}},"canonical_sha256":"abac3897d7881d6beca82c09c072f6ec37b8fe170daca1cb26bfb28f3a8b1e1b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"abac3897d7881d6beca82c09c072f6ec37b8fe170daca1cb26bfb28f3a8b1e1b","first_computed_at":"2026-05-18T02:44:49.236278Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:49.236278Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"F1j6TLUB2CJ0GK6rKzrnY94/5YqAjK32OZBrAdEuxnO60NvfO8v2ewvRmL3Ih4rdy1uA6eY4Q1X4xyNtqJ1JDQ==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:49.236719Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13276","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b32f3be5d9e564c00e212866440ef42a8e49a8b7749465474d267a87d5b1f7ca","sha256:53e9d9a7313a72e79995c12af06e32f082265cab9fea6262cc5689cbede5ffb9"],"state_sha256":"6837b37d9ecae33e33762e509afb5e20ab57f3e896238658242911305868c0be"}