PulseAugur
中
实时 19:55:29
English(EN) BOSS: Token-Budgeted Online Sampling and Selection for LLM Reasoning

新的BOSS策略在令牌预算下优化LLM推理

研究人员开发了BOSS,这是一种新颖的、无需训练的策略,旨在严格的令牌预算内优化大型语言模型(LLM)的推理。BOSS通过实时决定何时停止生成、生成另一个候选或评估现有候选来解决实例内分配的挑战。在Llama-3.1模型上跨四个推理基准进行了测试,BOSS在22个模型-任务-预算配置中表现出优于现有方法(如SDN中的自适应一致性多控制器策略和TALE-EP)的准确性。 AI

影响 这项新策略有望在计算约束严格的应用中提高LLM推理的效率和准确性。

排序理由 该条目描述了一篇关于LLM推理新策略的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

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

新的BOSS策略在令牌预算下优化LLM推理

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
该条目描述了一篇关于LLM推理新策略的最新研究论文。[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, model release
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
34 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]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    BOSS:LLM推理的令牌预算在线采样与选择

    Test-time scaling can improve LLM reasoning, but existing approaches typically adapt generation or apply answer selection after a predetermined generation schedule. We study within-instance allocation: under a hard per-instance prompt-plus-completion token cap, when should infere…