PulseAugur
实时 09:31:59
English(EN) GIFT: Reconciling Post-Training Objectives via Variational Finite-Temperature Gibbs Initialization

新的GIFT方法通过协调SFT和RL来增强大型推理模型训练

研究人员推出了一种新方法GIFT(Gibbs Initialization with Finite Temperature,有限温度吉布斯初始化),以改进大型推理模型(LRMs)的训练后过程。该技术通过使SFT目标在结构上与后续的RL阶段兼容,解决了监督微调(SFT)和强化学习(RL)之间的优化不匹配问题。通过采用源自KL正则化RL的吉布斯最优值的令牌级变分代理,GIFT通过有限温度保留了结构多样性,与标准SFT中出现的分布坍塌不同。实验表明,在用作RL初始化时,GIFT的表现优于传统的SFT和其他基线。 AI

影响 这项新的初始化技术可以通过改善监督学习和强化学习阶段之间的协同作用,从而实现更强大、更多样化的大型推理模型。

排序理由 该集群包含一篇详细介绍大型推理模型新训练方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 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GIFT方法通过协调SFT和RL来增强大型推理模型训练

本文如何被排名

Signal score
13 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群包含一篇详细介绍大型推理模型新训练方法的论文。[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
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准

报道来源 [1]

  1. arXiv cs.AI TIER_1 English(EN) · Zhengyang Zhao, Lu Ma, Yizhen Jiang, Xiaochen Ma, Zimo Meng, Chengyu Shen, Lexiang Tang, Haoze Sun, Peng Pei, Wentao Zhang ·

    GIFT:通过变分有限温度吉布斯初始化协调训练后目标

    arXiv:2601.09233v3 Announce Type: replace-cross Abstract: The prevailing post-training paradigm for Large Reasoning Models (LRMs)---Supervised Fine-Tuning (SFT) followed by Reinforcement Learning (RL)-suffers from an intrinsic optimization mismatch: the rigid likelihood maximizat…