Skip to content
AI News HubLIVE
Original source2 min read

Resource-Efficient Reinforcement Learning Post-Training through Elastic Expert Parallelism

Summary

FlexFloor is a system for reinforcement learning post-training of mixture-of-experts models that jointly adapts response scheduling and the active worker set, releasing workers earlier during synchronous response generation while preserving the expert computation still needed by ongoing responses. Against the verl baseline it produces roughly 65.40% more response tokens per unit of occupied compute time and raises response generation throughput by 52.74%, with little change in rollout reward.

SourcearXiv AIAuthor: Jia Wei, Ziyi Qiu, Yi Xu, Xuehai Qian
Resource-Efficient Reinforcement Learning Post-Training through Elastic Expert Parallelism
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 4 Nov 2026]

Title:Resource-Efficient Reinforcement Learning Post-Training through Elastic Expert Parallelism

View a PDF of the paper titled Resource-Efficient Reinforcement Learning Post-Training through Elastic Expert Parallelism, by Jia Wei and 3 other authors

View PDF HTML (experimental)

Abstract:Reinforcement learning post-training improves the capabilities of large language models, but generating training responses is a major computational bottleneck. In mixture-of-experts models, workers that finish their assigned responses may remain occupied because their experts are still needed by workers processing longer responses. Consequently, approximately one-third of occupied worker time in the baseline occurs after local response generation finishes. Here we present FlexFloor, a system that enables earlier resource release during synchronous response generation while preserving the expert computation required by ongoing responses. FlexFloor jointly adapts response scheduling and the active group of workers. It uses historical response lengths and memory constraints to guide batching and prompt placement, then progressively releases workers while maintaining expert availability and correct communication. Compared with the verl baseline, FlexFloor generates approximately 65.40% more response tokens per unit of occupied computing time and increases response generation throughput by 52.74%, with little observed variation in rollout reward. By making reinforcement learning post-training more efficient, FlexFloor supports faster development and iterative improvement of large language models and AI agents.

Comments: 19 pages including references and appendices, 5 figures

Subjects:

Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)

Cite as: arXiv:2611.06984 [cs.AI]

(or arXiv:2611.06984v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2611.06984

arXiv-issued DOI via DataCite

Submission history

From: Jia Wei [view email] [v1] Wed, 4 Nov 2026 15:36:14 UTC (1,055 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Resource-Efficient Reinforcement Learning Post-Training through Elastic Expert Parallelism, by Jia Wei and 3 other authors

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.AI

new | recent | 2026-11

Change to browse by:

cs cs.DC

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • The baseline wastes about one-third of occupied worker time after local response generation has finished
  • Jointly adapts response scheduling and the active worker group, progressively releasing workers while keeping experts available and communication correct
  • About 65.40% more response tokens per unit of occupied compute time and 52.74% higher throughput, with little variation in rollout reward

Highlights and analysis are generated automatically and may contain errors. Check the original source.

More recent stories on this topic · Models

Resource-Efficient Reinforcement Learning Post-Training through Elastic Expert Parallelism | AI News Hub