语言大模型 / LLM - arXivDaily 专题

2406.10892 2025-01-03 cs.LG 79%

DIPPER: Direct Preference Optimization to Accelerate Primitive-Enabled Hierarchical Reinforcement Learning

Utsav Singh, Souradip Chakraborty, Wesley A. Suttle, Brian M. Sadler, Vinay P Namboodiri, Amrit Singh Bedi

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.LG

Comments This paper is subsumed by a later paper of ours: arXiv:2411.00361

2412.19005 2024-12-30 eess.AS cs.AI 79%

Enhancing Audiovisual Speech Recognition through Bifocal Preference Optimization

Yihan Wu, Yichen Lu, Yifan Peng, Xihua Wang, Ruihua Song, Shinji Watanabe

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.AI

Comments Accepted by AAAI 2025

2412.17572 2024-12-24 cs.CV cs.AI 79%

Empathetic Response in Audio-Visual Conversations Using Emotion Preference Optimization and MambaCompressor

Yeonju Kim, Se Jin Park, Yong Man Ro

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.AI

2409.05622 2024-12-17 cs.LG 79%

Forward KL Regularized Preference Optimization for Aligning Diffusion Policies

Zhao Shan, Chenyou Fan, Shuang Qiu, Jiyuan Shi, Chenjia Bai

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.LG

2412.09899 2024-12-16 cs.LG 79%

TTAQ: Towards Stable Post-training Quantization in Continuous Domain Adaptation

Junrui Xiao, Zhikai Li, Lianwei Yang, Yiduo Mei, Qingyi Gu

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

2412.07268 2024-12-11 cs.LG cs.MM 79%

PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models

Zining Wnag, Jinyang Guo, Ruihao Gong, Yang Yong, Aishan Liu, Yushi Huang, Jiaheng Liu, Xianglong Liu

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

2407.14503 2024-11-11 cs.LG 79%

Catastrophic Goodhart: regularizing RLHF with KL divergence does not mitigate heavy-tailed reward misspecification

Thomas Kwa, Drake Thomas, Adrià Garriga-Alonso

专题命中后训练与偏好优化：RLHF(title,abstract)；分类 cs.LG

Comments Mechanistic Interpretability workshop at ICML 2024; Main conference poster at NeurIPS 2024

2405.14751 2024-11-06 cs.LG 79%

AGILE: A Novel Reinforcement Learning Framework of LLM Agents

Peiyuan Feng, Yichen He, Guanhua Huang, Yuan Lin, Hanchong Zhang, Yuchen Zhang, Hang Li

专题命中后训练与偏好优化：LLM(title,abstract)；分类 cs.LG

Comments accepted by NeurIPS 2024

2406.10203 2024-10-29 cs.CL 79%

A Probability--Quality Trade-off in Aligned Language Models and its Relation to Sampling Adaptors

Naaman Tan, Josef Valvoda, Tianyu Liu, Anej Svete, Yanxia Qin, Kan Min-Yen, Ryan Cotterell

专题命中后训练与偏好优化：language model(title,abstract)；分类 cs.CL

Comments EMNLP 2024

2406.18853 2024-10-29 cs.LG 79%

Decoding-Time Language Model Alignment with Multiple Objectives

Ruizhe Shi, Yifang Chen, Yushi Hu, Alisa Liu, Hannaneh Hajishirzi, Noah A. Smith, Simon S. Du

专题命中后训练与偏好优化：language model(title,abstract)；分类 cs.LG

Comments NeurIPS accepted version

2403.16576 2024-10-29 q-bio.BM cs.LG 79%

Antigen-Specific Antibody Design via Direct Energy-based Preference Optimization

Xiangxin Zhou, Dongyu Xue, Ruizhe Chen, Zaixiang Zheng, Liang Wang, Quanquan Gu

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.LG

Comments Accepted to NeurIPS 2024

2309.01829 2024-10-18 quant-ph cs.LG 79%

A Post-Training Approach for Mitigating Overfitting in Quantum Convolutional Neural Networks

Aakash Ravindra Shinde, Charu Jain, Amir Kalev

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

Comments 9 pages, 14 images, 6 tables

Journal ref Phys Rev A 110 042409 (2024)

2309.06256 2024-10-15 cs.LG 79%

Mitigating the Alignment Tax of RLHF

Yong Lin, Hangyu Lin, Wei Xiong, Shizhe Diao, Jianmeng Liu, Jipeng Zhang, Rui Pan, Haoxiang Wang, Wenbin Hu, Hanning Zhang, Hanze Dong, Renjie Pi, Han Zhao, Nan Jiang, Heng Ji, Yuan Yao, Tong Zhang

专题命中后训练与偏好优化：RLHF(title,abstract)；分类 cs.LG

Comments EMNLP 2024 Main

2407.14726 2024-07-30 cs.CV cs.LG 79%

MetaAug: Meta-Data Augmentation for Post-Training Quantization

Cuong Pham, Hoang Anh Dung, Cuong C. Nguyen, Trung Le, Dinh Phung, Gustavo Carneiro, Thanh-Toan Do

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

Comments Accepted by ECCV 2024

2406.11581 2024-07-30 cs.CL 79%

Style Transfer with Multi-iteration Preference Optimization

Shuai Liu, Jonathan May

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.CL

2311.06322 2024-07-09 cs.CV cs.LG 79%

Post-training Quantization for Text-to-Image Diffusion Models with Progressive Calibration and Activation Relaxing

Siao Tang, Xin Wang, Hong Chen, Chaoyu Guan, Zewen Wu, Yansong Tang, Wenwu Zhu

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

Comments Accepted by ECCV2024

2402.06886 2024-06-04 cs.LG math.OC stat.ML 79%

Principled Penalty-based Methods for Bilevel Reinforcement Learning and RLHF

Han Shen, Zhuoran Yang, Tianyi Chen

专题命中后训练与偏好优化：RLHF(title,abstract)；分类 cs.LG

Comments Shorter version accepted to ICML 2024

2405.18810 2024-05-30 cs.CV cs.AI 79%

UniPTS: A Unified Framework for Proficient Post-Training Sparsity

Jingjing Xie, Yuxin Zhang, Mingbao Lin, Zhihang Lin, Liujuan Cao, Rongrong Ji

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.AI

Comments Accepted by CVPR2024

2404.13038 2024-04-22 cs.AI cs.CY 79%

Mapping Social Choice Theory to RLHF

Jessica Dai, Eve Fleisig

专题命中后训练与偏好优化：RLHF(title,abstract)；分类 cs.AI

2401.06838 2024-04-16 cs.CL 79%

MAPO: Advancing Multilingual Reasoning through Multilingual Alignment-as-Preference Optimization

Shuaijie She, Wei Zou, Shujian Huang, Wenhao Zhu, Xiang Liu, Xiang Geng, Jiajun Chen

专题命中后训练与偏好优化：preference optimization(title,abstract)；分类 cs.CL

Comments The project is available at https://github.com/NJUNLP/MAPO

2404.01730 2024-04-03 cs.LG cs.IT math.IT stat.ML 79%

Asymptotics of Language Model Alignment

Joy Qiping Yang, Salman Salamatian, Ziteng Sun, Ananda Theertha Suresh, Ahmad Beirami

专题命中后训练与偏好优化：language model(title,abstract)；分类 cs.LG

2403.17031 2024-03-27 cs.LG 79%

The N+ Implementation Details of RLHF with PPO: A Case Study on TL;DR Summarization

Shengyi Huang, Michael Noukhovitch, Arian Hosseini, Kashif Rasul, Weixun Wang, Lewis Tunstall

专题命中后训练与偏好优化：RLHF(title,abstract)；分类 cs.LG

2402.18950 2024-03-01 cs.CL 79%

PopALM: Popularity-Aligned Language Models for Social Media Trendy Response Prediction

Erxin Yu, Jing Li, Chunpu Xu

专题命中后训练与偏好优化：language model(title,abstract)；分类 cs.CL

Comments Accepted by COLING 2024

2312.10584 2024-02-27 cs.LG 79%

Policy Optimization in RLHF: The Impact of Out-of-preference Data

Ziniu Li, Tian Xu, Yang Yu

专题命中后训练与偏好优化：RLHF(title)；preference optimization(abstract)；分类 cs.LG

2309.16827 2023-10-02 cs.LG 79%

Post-Training Overfitting Mitigation in DNN Classifiers

Hang Wang, David J. Miller, George Kesidis

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

2308.04332 2023-08-09 cs.LG cs.HC 79%

RLHF-Blender: A Configurable Interactive Interface for Learning from Diverse Human Feedback

Yannick Metz, David Lindner, Raphaël Baur, Daniel Keim, Mennatallah El-Assady

专题命中后训练与偏好优化：RLHF(title,abstract)；分类 cs.LG

Comments 14 pages, 3 figures

Journal ref ICML2023 Interactive Learning from Implicit Human Feedback Workshop

2308.04269 2023-08-09 cs.CV cs.AI 79%

Lossy and Lossless (L$^2$) Post-training Model Size Compression

Yumeng Shi, Shihao Bai, Xiuying Wei, Ruihao Gong, Jianlei Yang

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.AI

2205.06900 2023-08-08 cs.LG cs.CR 79%

MM-BD: Post-Training Detection of Backdoor Attacks with Arbitrary Backdoor Pattern Types Using a Maximum Margin Statistic

Hang Wang, Zhen Xiang, David J. Miller, George Kesidis

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.LG

2307.00569 2023-07-04 cs.CL 79%

SSP: Self-Supervised Post-training for Conversational Search

Quan Tu, Shen Gao, Xiaolong Wu, Zhao Cao, Ji-Rong Wen, Rui Yan

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.CL

Comments Accepted by ACL 2023 Findings, Long Paper

2106.07217 2023-04-20 cs.CV cs.AI 79%

Influential Rank: A New Perspective of Post-training for Robust Model against Noisy Labels

Seulki Park, Hwanjun Song, Daeho Um, Dae Ung Jo, Sangdoo Yun, Jin Young Choi

专题命中后训练与偏好优化：post-training(title,abstract)；分类 cs.AI

Comments 15 pages

1. 后训练与偏好优化 4541 篇

DIPPER: Direct Preference Optimization to Accelerate Primitive-Enabled Hierarchical Reinforcement Learning

Enhancing Audiovisual Speech Recognition through Bifocal Preference Optimization

Empathetic Response in Audio-Visual Conversations Using Emotion Preference Optimization and MambaCompressor

Forward KL Regularized Preference Optimization for Aligning Diffusion Policies

TTAQ: Towards Stable Post-training Quantization in Continuous Domain Adaptation

PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models

Catastrophic Goodhart: regularizing RLHF with KL divergence does not mitigate heavy-tailed reward misspecification

AGILE: A Novel Reinforcement Learning Framework of LLM Agents

A Probability--Quality Trade-off in Aligned Language Models and its Relation to Sampling Adaptors

Decoding-Time Language Model Alignment with Multiple Objectives

Antigen-Specific Antibody Design via Direct Energy-based Preference Optimization

A Post-Training Approach for Mitigating Overfitting in Quantum Convolutional Neural Networks

Mitigating the Alignment Tax of RLHF

MetaAug: Meta-Data Augmentation for Post-Training Quantization

Style Transfer with Multi-iteration Preference Optimization

Post-training Quantization for Text-to-Image Diffusion Models with Progressive Calibration and Activation Relaxing

Principled Penalty-based Methods for Bilevel Reinforcement Learning and RLHF

UniPTS: A Unified Framework for Proficient Post-Training Sparsity

Mapping Social Choice Theory to RLHF

MAPO: Advancing Multilingual Reasoning through Multilingual Alignment-as-Preference Optimization

Asymptotics of Language Model Alignment

The N+ Implementation Details of RLHF with PPO: A Case Study on TL;DR Summarization

PopALM: Popularity-Aligned Language Models for Social Media Trendy Response Prediction

Policy Optimization in RLHF: The Impact of Out-of-preference Data

Post-Training Overfitting Mitigation in DNN Classifiers

RLHF-Blender: A Configurable Interactive Interface for Learning from Diverse Human Feedback

Lossy and Lossless (L$^2$) Post-training Model Size Compression

MM-BD: Post-Training Detection of Backdoor Attacks with Arbitrary Backdoor Pattern Types Using a Maximum Margin Statistic

SSP: Self-Supervised Post-training for Conversational Search

Influential Rank: A New Perspective of Post-training for Robust Model against Noisy Labels