# Reflex reuses and corrects past robot actions to speed policy serving [Canonical reading](https://escapemeridian.com/readings/reflex-reuses-and-corrects-past-robot-actions-to-speed-policy-0c982c49) In the arXiv abstract, authors Ziyang Lin and coauthors argue that speeding up robot policy serving requires more than reusing intermediate computations. They say that this reuse can offer little advantage over running fewer denoising steps, while the policy backbone can still account for 70% of GPU time. Their proposed serving layer, Reflex, caches complete action chunks from past policy executions, bypassing both the backbone and action head. Because similar observations can call for different actions, and reuse errors can accumulate into execution stalls, Reflex uses action-level correction and refreshes the original policy based on task progress. The authors report inference latency improvements of 1.4x to 11x and throughput improvements of 2.0x to 2.2x on representative workloads and real robots, while preserving comparable task success. These are reported results, not evidence here of performance across other robots or tasks. The abstract does not specify the workloads, task-success measurements, or how often the original policy must be refreshed. How well does Reflex preserve task success across different tasks and robots as it reuses cached action chunks? Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims. ## Original source arXiv cs.AI · Article Original source date: Oct 31, 2026 (2026-10-31). Date basis: Date basis not provided. Added to Escape Meridian: 2026-11-10T06:31:27.970Z. First observed: 2026-11-03T09:04:04.314Z. Retrieved: 2026-11-04T01:33:11.522Z. - [Reflex: Fast Robot Policy Serving by Reusing and Correcting Past Actions](https://arxiv.org/abs/2611.00524) ## Limits Evidence is the arXiv abstract; it does not provide details of the representative workloads, task-success measurements, or policy refresh frequency. The abstract's reported results do not establish performance across other tasks or robots. ## Open question How well does Reflex preserve task success across different tasks and robots as it reuses cached action chunks? ## Source-stated The authors say reusing intermediate denoising states can offer little advantage over running fewer denoising steps from fresh noise, and that the policy backbone can account for 70% of GPU time. > such reuse often offers little advantage over simply running fewer denoising steps from fresh noise, without any cache. Moreover, accelerating denoising leaves the expensive policy backbone intact, which can account for 70% of GPU time. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source-stated Reflex caches complete action chunks from past policy executions to bypass the backbone and action head. > Reflex caches complete action chunks to bypass both the backbone and action head. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source-stated Reflex uses action-level correction and task-progress-based policy refresh to address action mismatch and accumulated errors. > At the execution level, it captures task progress and adaptively refreshes the original policy to limit accumulated errors and balance inference cost. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source-stated The authors report inference latency improvements of 1.4x to 11x and throughput improvements of 2.0x to 2.2x, with comparable task success on representative workloads and real robots. > Reflex improves inference latency by 1.4x to 11x and throughput by 2.0x to 2.2x on representative workloads and real robots, while preserving comparable task success. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source and contributor context Lin, Ziyang, Sun, Zixuan, Nasser, Rami, Jin, Jianli, Li, Xuanle, Srinivasa, Jayanth, Lai, Fan · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media No suitable media was found by the supported page extraction. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Graphics processing unit A chip that performs many calculations at once; the main hardware used to train and run AI models. - [Definition reference](https://en.wikipedia.org/wiki/Graphics_processing_unit) ## Latency The delay between sending a request and receiving the result. - [Definition reference](https://developers.google.com/machine-learning/glossary#latency) ## arXiv arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed. - [Explore this context](/entities/arxiv) - [Reference](https://arxiv.org/) - [Reference](https://info.arxiv.org/help/api/index.html) ## TIME TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services. - [Explore this context](/entities/time) - [Reference](https://time.com/) ## Editorial provenance Version 1. Publication ID: 50198add-57fe-4212-8cab-657f2473685f. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-10T06:31:27.970Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: material\_development. Observation: 8c687d7142571d063d16da3919a9c5a99a066acdb84943363834461a15cac3c9. Source text hash: 940f4a9fa26013ec77dc29a2a2f869e2508c52215156acf68dba3fc0d878262b. ## Revision and source-check history ### Version 1 — 2026-11-10T06:31:27.970Z In the arXiv abstract, authors Ziyang Lin and coauthors argue that speeding up robot policy serving requires more than reusing intermediate computations. They say that this reuse can offer little advantage over running fewer denoising steps, while the policy backbone can still account for 70% of GPU time. Their proposed serving layer, Reflex, caches complete action chunks from past policy executions, bypassing both the backbone and action head. Because similar observations can call for different actions, and reuse errors can accumulate into execution stalls, Reflex uses action-level correction and refreshes the original policy based on task progress. The authors report inference latency improvements of 1.4x to 11x and throughput improvements of 2.0x to 2.2x on representative workloads and real robots, while preserving comparable task success. These are reported results, not evidence here of performance across other robots or tasks. The abstract does not specify the workloads, task-success measurements, or how often the original policy must be refreshed. How well does Reflex preserve task success across different tasks and robots as it reuses cached action chunks? ## Original source arXiv cs.AI · Article Original source date: Oct 31, 2026 (2026-10-31). Date basis: Date basis not provided. Added to Escape Meridian: 2026-11-10T06:31:27.970Z. First observed: 2026-11-03T09:04:04.314Z. Retrieved: 2026-11-04T01:33:11.522Z. - [Reflex: Fast Robot Policy Serving by Reusing and Correcting Past Actions](https://arxiv.org/abs/2611.00524) ## Limits Evidence is the arXiv abstract; it does not provide details of the representative workloads, task-success measurements, or policy refresh frequency. The abstract's reported results do not establish performance across other tasks or robots. ## Open question How well does Reflex preserve task success across different tasks and robots as it reuses cached action chunks? ## Source-stated The authors say reusing intermediate denoising states can offer little advantage over running fewer denoising steps from fresh noise, and that the policy backbone can account for 70% of GPU time. > such reuse often offers little advantage over simply running fewer denoising steps from fresh noise, without any cache. Moreover, accelerating denoising leaves the expensive policy backbone intact, which can account for 70% of GPU time. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source-stated Reflex caches complete action chunks from past policy executions to bypass the backbone and action head. > Reflex caches complete action chunks to bypass both the backbone and action head. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source-stated Reflex uses action-level correction and task-progress-based policy refresh to address action mismatch and accumulated errors. > At the execution level, it captures task progress and adaptively refreshes the original policy to limit accumulated errors and balance inference cost. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source-stated The authors report inference latency improvements of 1.4x to 11x and throughput improvements of 2.0x to 2.2x, with comparable task success on representative workloads and real robots. > Reflex improves inference latency by 1.4x to 11x and throughput by 2.0x to 2.2x on representative workloads and real robots, while preserving comparable task success. - [Supporting source](https://arxiv.org/abs/2611.00524) ## Source and contributor context Lin, Ziyang, Sun, Zixuan, Nasser, Rami, Jin, Jianli, Li, Xuanle, Srinivasa, Jayanth, Lai, Fan · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media No suitable media was found by the supported page extraction. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Graphics processing unit A chip that performs many calculations at once; the main hardware used to train and run AI models. - [Definition reference](https://en.wikipedia.org/wiki/Graphics_processing_unit) ## Latency The delay between sending a request and receiving the result. - [Definition reference](https://developers.google.com/machine-learning/glossary#latency) ## arXiv arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed. - [Explore this context](/entities/arxiv) - [Reference](https://arxiv.org/) - [Reference](https://info.arxiv.org/help/api/index.html) ## TIME TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services. - [Explore this context](/entities/time) - [Reference](https://time.com/) ## Editorial provenance Version 1. Publication ID: 50198add-57fe-4212-8cab-657f2473685f. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-10T06:31:27.970Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: material\_development. Observation: 8c687d7142571d063d16da3919a9c5a99a066acdb84943363834461a15cac3c9. Source text hash: 940f4a9fa26013ec77dc29a2a2f869e2508c52215156acf68dba3fc0d878262b. ### uncertain — 2026-11-10T06:31:27.970Z — anthropic/claude-opus-5.5 Publication: 50198add-57fe-4212-8cab-657f2473685f attribution — pass: The draft attributes the argument and the reported results to the paper's authors, led by Ziyang Lin, in an arXiv abstract. Speedup and success figures are framed as the authors' reports, not established facts. > Authors:Ziyang Lin, Zixuan Sun, Rami Nasser, Jianli Jin, Xuanle Li, Jayanth Srinivasa, Fan Lai summary — uncertain: The draft is accurate, but a non-specialist may struggle to follow it, and it omits the stakes. It never says why serving speed matters: the source names latency as a bottleneck for real-time robot control and throughput as a bottleneck for multi-robot serving. Central terms are not explained in plain words or flagged for tooltips: robot policy, denoising steps, backbone, action head, action chunks, latency and throughput. It also omits the condition that Reflex works without modifying the underlying policy. Smallest repair: add the source's real-time and multi-robot motivation, add the no-modification condition, and make sure these terms are explained. > making serving a latency bottleneck for real-time control and a throughput bottleneck for multi-robot serving date — pass: The originalPublished date, 2026-10-31, matches the arXiv v1 submission date. It is not the later feed date of 2026-11-03. > \[Submitted on 31 Oct 2026\] support — pass: All claims track the abstract: the limited benefit of reusing denoising states, the 70% backbone share of GPU time, caching full action chunks, two-level correction with policy refresh, and the 1.4x to 11x latency and 2.0x to 2.2x throughput gains with comparable task success. The draft's caveats about scope and undisclosed details are appropriate. Rendering the source's 'often offers little advantage' as 'can offer' is a minor softening, not a contradiction. > Reflex improves inference latency by 1.4x to 11x and throughput by 2.0x to 2.2x on representative workloads and real robots, while preserving comparable task success. media — pass: The selected media is the verified direct link to the original arXiv abstract page. It matches the source, and no better media was found. ## Recorded references Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter. No public references recorded here yet. [Reference data (JSON)](/readings/0c982c49462dbf1168f4ec0e35b505e7dba0ae4894dfee667dcacf06290177d2/references.json) ## Corrections and challenges [Send a correction or an opposing source](https://escapemeridian.com/corrections/) [JSON record](https://escapemeridian.com/readings/reflex-reuses-and-corrects-past-robot-actions-to-speed-policy-0c982c49.json)