Skip to content
← Back to explorer

Periodic Asynchrony: An On-Policy Approach for Accelerating LLM Reinforcement Learning

Jian Lu · Nov 24, 2025 · Citations: 0

Open arXiv Find Implementation RSS feed Shortlist (0)

Data freshness

Extraction: Stale

Check recency before relying on this page for active eval decisions. Use stale pages as context and verify against current hub results.

Metadata refreshed

Mar 10, 2026, 1:59 PM

Stale

Extraction refreshed

Mar 10, 2026, 1:59 PM

Stale

Extraction source

Persisted extraction

Confidence unavailable

Abstract

Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge. In mainstream RL frameworks, inference and training are co-located on the same devices, and their synchronous execution prevents concurrent inference and training. In this work, we revisit the strategy of separating inference and training deployment, and propose a periodically asynchronous framework that transforms synchronous RL training into an asynchronous producer-consumer pipeline. Unlike existing asynchronous approaches that introduce off-policy bias, our design is provably equivalent to its synchronous counterpart, preserving strict on-policy correctness without any algorithmic modifications. We further introduce a unified tri-model architecture and a shared-prompt attention mechanism to support efficient asynchronous execution and reduce redundant computation. Experiments on NPU platforms demonstrate a three- to five-fold improvement in end-to-end training throughput over mainstream RL frameworks, while maintaining fully comparable accuracy, indicating its potential for widespread application.

Low-signal caution for protocol decisions

Use this page for context, then validate protocol choices against stronger HFEPX references before implementation decisions.

  • Structured extraction is still processing; current fields are metadata-first.
Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub

HFEPX Relevance Assessment

Signal extraction is still processing. This page currently shows metadata-first guidance until structured protocol fields are ready.

Best use

Background context only

Use if you need

A provisional background reference while structured extraction finishes.

Main weakness

Structured extraction is still processing; current fields are metadata-first.

Trust level

Provisional

Eval-Fit Score

Unavailable

Eval-fit score is unavailable until extraction completes.

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Weak / implicit signal

HFEPX Fit

Provisional (processing)

Extraction confidence: Provisional

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

Field Provenance & Confidence

Each key protocol field shows extraction state, confidence band, and data source so you can decide whether to trust it directly or validate from full text.

Human Feedback Types

provisional

None explicit

Confidence: Provisional Source: Persisted extraction inferred

No explicit feedback protocol extracted.

Evidence snippet: Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.

Evaluation Modes

provisional

Automatic metrics

Confidence: Provisional Source: Persisted extraction inferred

Includes extracted eval setup.

Evidence snippet: Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.

Quality Controls

provisional

Not reported

Confidence: Provisional Source: Persisted extraction inferred

No explicit QC controls found.

Evidence snippet: Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.

Benchmarks / Datasets

provisional

Not extracted

Confidence: Provisional Source: Persisted extraction inferred

No benchmark anchors detected.

Evidence snippet: Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.

Reported Metrics

provisional

Accuracy

Confidence: Provisional Source: Persisted extraction inferred

Useful for evaluation criteria comparison.

Evidence snippet: Experiments on NPU platforms demonstrate a three- to five-fold improvement in end-to-end training throughput over mainstream RL frameworks, while maintaining fully comparable accuracy, indicating its potential for widespread application.

Rater Population

provisional

Unknown

Confidence: Provisional Source: Persisted extraction inferred

Rater source not explicitly reported.

Evidence snippet: Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.

Human Data Lens

Structured extraction is still processing. Below are provisional signals inferred from abstract text only.

  • Potential human-data signal: No explicit human-data keywords detected.
  • Potential benchmark anchors: No benchmark names detected in abstract.
  • Abstract highlights: 3 key sentence(s) extracted below.

Evaluation Lens

Evaluation fields are currently inferred heuristically from abstract text.

  • Potential evaluation modes: Automatic metrics
  • Potential metric signals: Accuracy
  • Confidence: Provisional (metadata-only fallback).

Research Brief

Deterministic synthesis

Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.

Generated Mar 10, 2026, 1:59 PM · Grounded in abstract + metadata only

Key Takeaways

  • Since the introduction of the GRPO algorithm, reinforcement learning (RL) has attracted increasing attention for LLM post-training, yet training efficiency remains a critical challenge.
  • In mainstream RL frameworks, inference and training are co-located on the same devices, and their synchronous execution prevents concurrent inference and training.
  • In this work, we revisit the strategy of separating inference and training deployment, and propose a periodically asynchronous framework that transforms synchronous RL training into an asynchronous producer-consumer pipeline.

Researcher Actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Validate inferred eval signals (Automatic metrics) against the full paper.
  • Use related-paper links to find stronger protocol-specific references.

Caveats

  • Generated from abstract + metadata only; no PDF parsing.
  • Signals below are heuristic and may miss details reported outside the abstract.

Recommended Queries

Accuracy

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.

No related papers found for this item yet.

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote