English(EN) Deterministic Inference across Tensor Parallel Sizes That Eliminates Training-Inference Mismatch

新内核确保跨张量并行大小的确定性LLM推理

作者 PulseAugur 编辑部 · [1 个来源] · 2026-06-01 04:00

研究人员开发了基于树的不变内核（TBIK），以确保大型语言模型（LLM）的确定性推理，无论张量并行（TP）大小如何。这解决了由于TP大小和浮点算术的差异导致相同输入产生不同输出的关键问题。TBIK通过分层二叉树结构对齐归约顺序，保证了逐位可复现性，这对于LLM作为裁判和强化学习等应用至关重要。 AI

影响确保RL和评估等关键应用的LLM输出一致，消除了可靠部署的一个主要障碍。

排序理由该集群包含一篇学术论文，详细介绍了改进LLM推理的新技术方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv stat.ML 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇学术论文，详细介绍了改进LLM推理的新技术方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 113 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv stat.ML TIER_1 English(EN) · Ziyang Zhang, Xinheng Ding, Jiayi Yuan, Rixin Liu, Huizi Mao, Jiarong Xing, Zirui Liu · 2026-06-01 04:00

消除训练-推理不匹配的跨张量并行大小的确定性推理

arXiv:2511.17826v2 Announce Type: replace-cross Abstract: Deterministic inference is increasingly critical for large language model (LLM) applications such as LLM-as-a-judge evaluation, multi-agent systems, and Reinforcement Learning (RL). However, existing LLM serving frameworks…

本文如何被排名

报道来源 [1]

消除训练-推理不匹配的跨张量并行大小的确定性推理

相关实体

相关话题