中文
相关论文

相关论文: Beyond Majority Voting: Towards Fine-grained and M…

200 篇论文

Semi-supervised learning leverages unlabeled data to enhance model performance, addressing the limitations of fully supervised approaches. Among its strategies, pseudo-supervision has proven highly effective, typically relying on one or…

计算机视觉与模式识别 · 计算机科学 2025-05-13 Negin Ghamsarian , Sahar Nasirihaghighi , Klaus Schoeffmann , Raphael Sznitman

Reinforcement Learning with Verifiable Rewards (RLVR) elicits long chain-of-thought reasoning in large language models (LLMs), but outcome-based rewards lead to coarse-grained advantage estimation. While existing approaches improve RLVR via…

计算与语言 · 计算机科学 2026-01-08 Fei Wu , Zhenrong Zhang , Qikai Chang , Jianshu Zhang , Quan Liu , Jun Du

Semi-supervised learning is the problem of training an accurate predictive model by combining a small labeled dataset with a presumably much larger unlabeled dataset. Many methods for semi-supervised deep learning have been developed,…

机器学习 · 计算机科学 2023-10-31 Sumeet Menon , David Chapman

Preference-based reinforcement learning (RL) has shown potential for teaching agents to perform the target tasks without a costly, pre-defined reward function by learning the reward with a supervisor's preference between the two agent…

机器学习 · 计算机科学 2022-03-21 Jongjin Park , Younggyo Seo , Jinwoo Shin , Honglak Lee , Pieter Abbeel , Kimin Lee

Reinforcement learning with human-annotated data has boosted chain-of-thought reasoning in large reasoning models, but these gains come at high costs in labeled data while faltering on harder tasks. A natural next step is experience-driven…

计算与语言 · 计算机科学 2025-10-03 Zhaoning Yu , Will Su , Leitian Tao , Haozhu Wang , Aashu Singh , Hanchao Yu , Jianyu Wang , Hongyang Gao , Weizhe Yuan , Jason Weston , Ping Yu , Jing Xu

Test-Time Reinforcement Learning (TTRL) enables Large Language Models (LLMs) to enhance reasoning capabilities on unlabeled test streams by deriving pseudo-rewards from majority voting consensus. However, existing TTRL methods rely…

机器学习 · 计算机科学 2026-04-21 Dong Yan , Jian Liang , Yanbo Wang , Shuo Lu , Ran He , Tieniu Tan

Label-free reinforcement learning enables large language models to improve reasoning capabilities without ground-truth supervision, typically by treating majority-voted answers as pseudo-labels. However, we identify a critical failure mode:…

计算与语言 · 计算机科学 2026-03-24 Teng Pan , Yuchen Yan , Zixuan Wang , Ruiqing Zhang , Guiyang Hou , Wenqi Zhang , Weiming Lu , Jun Xiao , Yongliang Shen

Semi-supervised learning (SSL) has witnessed great progress with various improvements in the self-training framework with pseudo labeling. The main challenge is how to distinguish high-quality pseudo labels against the confirmation bias.…

机器学习 · 计算机科学 2024-02-21 Siyuan Li , Weiyang Jin , Zedong Wang , Fang Wu , Zicheng Liu , Cheng Tan , Stan Z. Li

Large Language Models (LLMs) have achieved remarkable advancements in reasoning capabilities empowered by Reinforcement Learning with Verifiable Rewards (RLVR). Nonetheless, RLVR intrinsically relies on ground-truth labels for reward…

机器学习 · 计算机科学 2026-05-26 Li Wang , Xiaodong Lu , Xiaohan Wang , Yikun Ban , Jiajun Chai , Wei Lin , Tianhao Peng , Guojun Yin

The rising cost of acquiring supervised data has driven significant interest in self-improvement for large language models (LLMs). Straightforward unsupervised signals like majority voting have proven effective in generating pseudo-labels…

计算与语言 · 计算机科学 2026-04-01 Chunyang Jiang , Yonggang Zhang , Yiyang Cai , Chi-Min Chan , Yulong Liu , Mingming Chen , Wei Xue , Yike Guo

Fine tuning self supervised pretrained models using pseudo labels can effectively improve speech recognition performance. But, low quality pseudo labels can misguide decision boundaries and degrade performance. We propose a simple yet…

声音 · 计算机科学 2022-11-01 Zezhong Jin , Dading Zhong , Xiao Song , Zhaoyi Liu , Naipeng Ye , Qingcheng Zeng

Fine-tuning pretrained ASR models for specific domains is challenging when labeled data is scarce. But unlabeled audio and labeled data from related domains are often available. We propose an incremental semi-supervised learning pipeline…

Weakly supervised text classification methods typically train a deep neural classifier based on pseudo-labels. The quality of pseudo-labels is crucial to final performance but they are inevitably noisy due to their heuristic nature, so…

计算与语言 · 计算机科学 2022-10-26 Dheeraj Mekala , Chengyu Dong , Jingbo Shang

Previous LLMs-based RL studies typically follow either supervised learning with high annotation costs, or unsupervised paradigms using voting or entropy-based rewards. However, their performance remains far from satisfactory due to the…

机器学习 · 计算机科学 2026-04-22 Zhiyin Yu , Bo Zhang , Qibin Hou , Zhonghai Wu , Xiao Luo , Lei Bai

We propose a novel semi-supervised learning (SSL) method that adopts selective training with pseudo labels. In our method, we generate hard pseudo-labels and also estimate their confidence, which represents how likely each pseudo-label is…

机器学习 · 计算机科学 2021-03-16 Masato Ishii

While semi-supervised learning (SSL) has yielded promising results, the more realistic SSL scenario remains to be explored, in which the unlabeled data exhibits extremely high recognition difficulty, e.g., fine-grained visual classification…

机器学习 · 计算机科学 2023-12-20 Yue Duan , Zhen Zhao , Lei Qi , Luping Zhou , Lei Wang , Yinghuan Shi

Reinforcement learning with verifiable rewards (RLVR) has proven effective in training large reasoning models (LRMs) by leveraging answer-verifiable signals to guide policy optimization, which, however, suffers from high annotation costs.…

Speech Emotion Recognition (SER) is typically trained and evaluated on majority-voted labels, which simplifies benchmarking but masks subjectivity and provides little transparency into why predictions are made. This neglects valid minority…

音频与语音处理 · 电气工程与系统科学 2026-02-06 Bo-Hao Su , Hui-Ying Shih , Jinchuan Tian , Jiatong Shi , Chi-Chun Lee , Carlos Busso , Shinji Watanabe

In class-incremental semantic segmentation, we have no access to the labeled data of previous tasks. Therefore, when incrementally learning new classes, deep neural networks suffer from catastrophic forgetting of previously learned…

计算机视觉与模式识别 · 计算机科学 2022-03-14 Lu Yu , Xialei Liu , Joost van de Weijer

Self-learning is a classical approach for learning with both labeled and unlabeled observations which consists in giving pseudo-labels to unlabeled training instances with a confidence score over a predetermined threshold. At the same time,…

机器学习 · 计算机科学 2021-09-30 Vasilii Feofanov , Emilie Devijver , Massih-Reza Amini
‹ 上一页 1 2 3 10 下一页 ›