Accelerating Long-Tail Generation in Synchronous RLHF Training via Adaptive Tensor Parallelism
The paper presents a new method called PAT for improving the efficiency of Reinforcement Learning from Human Feedback (RLHF) training. PAT employs adaptive tensor parallelism to dynamically adjust configurations during the generation stage, addressing issues of underutilization of GPUs. Evaluations show that this method significantly reduces generation latency and overall training iteration time compared to existing frameworks.
- ▪Reinforcement Learning from Human Feedback (RLHF) is a key post-training paradigm for enhancing model quality.
- ▪The proposed PAT method dynamically reconfigures tensor parallelism during RLHF training to improve efficiency.
- ▪Evaluations indicate that PAT can reduce generation latency by up to 34.6% and end-to-end training iteration latency by up to 27.2%.
arXiv cs.AI files mainly under ai research. We currently carry 1,128 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | arXiv cs.AI |
| Canonical URL | https://arxiv.org/abs/2605.23945 |
| Publication time | Tue, 26 May 2026 00:00:00 -0400 |
| Retrieval time | 2026-05-26T04:07:43.013Z |
| Last seen | 2026-05-26T04:07:43.013Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | xqd9w2Y6QfL- |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.23945 (cs) [Submitted on 3 May 2026] Title:Accelerating Long-Tail Generation in Synchronous RLHF Training via Adaptive Tensor Parallelism Authors:Long Zhao, Qinghe Wang, Jiaan Zhu, Youhui Bai, Zewen Jin, Chaoyi Ruan, Shengnan Wang, Cheng Li View a PDF of the paper titled Accelerating Long-Tail Generation in Synchronous RLHF Training via Adaptive Tensor Parallelism, by Long Zhao and 7 other authors View PDF HTML (experimental) Abstract:Reinforcement Learning from Human Feedback (RLHF) has become a key post-training paradigm for improving model quality.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
