Results 251 to 260 of about 2,403,325 (304)

Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models

International Conference on Machine Learning
Harnessing the power of human-annotated data through Supervised Fine-Tuning (SFT) is pivotal for advancing Large Language Models (LLMs). In this paper, we delve into the prospect of growing a strong LLM out of a weak one without the need for acquiring ...
Zixiang Chen, Yihe Deng, Huizhuo Yuan, Kaixuan Ji, Quanquan Gu   +4 more
semanticscholar   +1 more source

ConRFT: A Reinforced Fine-tuning Method for VLA Models via Consistency Policy

Robotics
Vision-Language-Action (VLA) models have shown substantial potential in real-world robotic manipulation. However, fine-tuning these models through supervised learning struggles to achieve robust performance due to limited, inconsistent demonstrations ...
Yuhui Chen, Shuai Tian, Shugao Liu, Yingting Zhou, Haoran Li, Dongbin Zhao   +5 more
semanticscholar   +1 more source

Optimizing Test-Time Compute via Meta Reinforcement Fine-Tuning

International Conference on Machine Learning
Training models to effectively use test-time compute is crucial for improving the reasoning performance of LLMs. Current methods mostly do so via fine-tuning on search traces or running RL with 0/1 outcome reward, but do these approaches efficiently ...
Yuxiao Qu, Matthew Y. R. Yang, Amrith Rajagopal Setlur, Lewis Tunstall, E. Beeching, R. Salakhutdinov, Aviral Kumar   +6 more
semanticscholar   +1 more source

VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning

arXiv.org
Reinforcement Learning (RL) benefits Large Language Models (LLMs) for complex reasoning. Inspired by this, we explore integrating spatio-temporal specific rewards into Multimodal Large Language Models (MLLMs) to address the unique challenges of video ...
Xinhao Li, Ziang Yan, Desen Meng, Lu Dong, Xiangyun Zeng, Yinan He, Yali Wang, Yu Qiao, Yi Wang, Limin Wang   +9 more
semanticscholar   +1 more source

SRFT: A Single-Stage Method with Supervised and Reinforcement Fine-Tuning for Reasoning

arXiv.org
Large language models (LLMs) have achieved remarkable progress in reasoning tasks, yet the optimal integration of Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) remains a fundamental challenge.
Yuqian Fu, Ting Chen, Jiajun Chai, Xihuai Wang, Songjun Tu, Guojun Yin, Wei Lin, Qichao Zhang, Yuanheng Zhu, Dongbin Zhao   +9 more
semanticscholar   +1 more source

ReinFlow: Fine-tuning Flow Matching Policy with Online Reinforcement Learning

Neural Information Processing Systems
We propose ReinFlow, a simple yet effective online reinforcement learning (RL) framework that fine-tunes a family of flow matching policies for continuous robotic control.
Tonghe Zhang, Chao Yu, Sichang Su, Yu Wang   +3 more
semanticscholar   +1 more source

Parameter-Efficient Fine-Tuning for Foundation Models

arXiv.org
This survey delves into the realm of Parameter-Efficient Fine-Tuning (PEFT) within the context of Foundation Models (FMs). PEFT, a cost-effective fine-tuning technique, minimizes parameters and computational complexity while striving for optimal ...
Dan Zhang, Tao Feng, Lilong Xue, Yuandong Wang, Yuxiao Dong, Jie Tang   +5 more
semanticscholar   +1 more source

Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models

International Conference on Machine Learning
Current vision large language models (VLLMs) exhibit remarkable capabilities yet are prone to generate harmful content and are vulnerable to even the simplest jailbreaking attacks.
Yongshuo Zong, Ondrej Bohdal, Tingyang Yu, Yongxin Yang, Timothy M. Hospedales   +4 more
semanticscholar   +1 more source

Aligning Modalities in Vision Large Language Models via Preference Fine-tuning

arXiv.org
Instruction-following Vision Large Language Models (VLLMs) have achieved significant progress recently on a variety of tasks. These approaches merge strong pre-trained vision models and large language models (LLMs).
Yiyang Zhou, Chenhang Cui, Rafael Rafailov, Chelsea Finn, Huaxiu Yao   +4 more
semanticscholar   +1 more source

Does Fine-Tuning LLMs on New Knowledge Encourage Hallucinations?

Conference on Empirical Methods in Natural Language Processing
When large language models are aligned via supervised fine-tuning, they may encounter new factual information that was not acquired through pre-training.
Zorik Gekhman, G. Yona, Roee Aharoni, Matan Eyal, Amir Feder, Roi Reichart, Jonathan Herzig   +6 more
semanticscholar   +1 more source