English
Related papers

Related papers: Implicit Compression Regularization: Concise Reaso…

200 papers

Reinforcement learning (RL) is central to improving reasoning in large language models (LLMs) but typically requires ground-truth rewards. Test-Time Reinforcement Learning (TTRL) removes this need by using majority-vote rewards, but relies…

Machine Learning · Computer Science 2025-10-06 Aleksei Arzhantsev , Otmane Sakhi , Flavian Vasile

Recent large reasoning models (LRMs) driven by reinforcement learning algorithms (e.g., GRPO) have achieved remarkable performance on challenging reasoning tasks. However, these models suffer from overthinking, generating unnecessarily long…

Artificial Intelligence · Computer Science 2026-03-03 Gang Li , Yan Chen , Ming Lin , Tianbao Yang

There is a belief that learning to compress well will lead to intelligence. Recently, language modeling has been shown to be equivalent to compression, which offers a compelling rationale for the success of large language models (LLMs): the…

Computation and Language · Computer Science 2024-08-20 Yuzhen Huang , Jinghan Zhang , Zifei Shan , Junxian He

Deep neural networks suffer from catastrophic forgetting, where performance on previous tasks degrades after training on a new task. This issue arises due to the model's tendency to overwrite previously acquired knowledge with new…

Machine Learning · Computer Science 2025-12-02 Lama Alssum , Hasan Abed Al Kader Hammoud , Motasem Alfarra , Juan C Leon Alcazar , Bernard Ghanem

The thinking-while-speaking paradigm aims to make AI communication more human. A key challenge is maintaining fluent speech while performing deep reasoning. Our method, InterRS, tackles this by inserting reasoning steps only during natural…

Computation and Language · Computer Science 2026-05-21 Xuan Du , Qiangyu Yan , Wenshuo Li , Borui Jiang , Changming Xiao , Han Shu , Xinghao Chen

The aim of Inverse Reinforcement Learning (IRL) is to infer a reward function $R$ from a policy $\pi$. To do this, we need a model of how $\pi$ relates to $R$. In the current literature, the most common models are optimality, Boltzmann…

Machine Learning · Computer Science 2023-03-27 Joar Skalse , Alessandro Abate

Compression methods, including quantization, distillation, and pruning, improve the computational efficiency of large reasoning models (LRMs). However, existing studies either fail to sufficiently compare all three compression methods on…

Machine Learning · Computer Science 2026-03-03 Nan Zhang , Eugene Kwek , Yusen Zhang , Ngoc-Hieu Nguyen , Prasenjit Mitra , Rui Zhang

Multimodal Large Language Models (MLLMs) are powerful at integrating diverse data, but they often struggle with complex reasoning. While Reinforcement learning (RL) can boost reasoning in LLMs, applying it to MLLMs is tricky. Common issues…

Machine Learning · Computer Science 2025-06-30 Minjie Hong , Zirun Guo , Yan Xia , Zehan Wang , Ziang Zhang , Tao Jin , Zhou Zhao

Lossy image compression networks aim to minimize the latent entropy of images while adhering to specific distortion constraints. However, optimizing the neural network can be challenging due to its nature of learning quantized latent…

Computer Vision and Pattern Recognition · Computer Science 2025-03-14 Yingwen Zhang , Meng Wang , Xihua Sheng , Peilin Chen , Junru Li , Li Zhang , Shiqi Wang

Large Reasoning Models (LRMs) have shown remarkable reasoning capabilities, yet they often suffer from overthinking, expending redundant computational steps on simple problems, or underthinking, failing to explore sufficient reasoning paths…

Artificial Intelligence · Computer Science 2026-04-03 Yulin Li , Tengyao Tu , Li Ding , Junjie Wang , Huiling Zhen , Yixin Chen , Yong Li , Zhuotao Tian

Inverse reinforcement learning (IRL) aims to learn a reward function and a corresponding policy that best fit the demonstrated trajectories of an expert. However, current IRL works cannot learn incrementally from an ongoing trajectory…

Machine Learning · Computer Science 2025-07-24 Shicheng Liu , Minghui Zhu

Reinforcement learning (RL) can align language models with non-differentiable reward signals, such as human preferences. However, a major challenge arises from the sparsity of these reward signals - typically, there is only a single reward…

Computation and Language · Computer Science 2024-02-20 Meng Cao , Lei Shu , Lei Yu , Yun Zhu , Nevan Wichers , Yinxiao Liu , Lei Meng

Group-Relative Policy Optimization (GRPO) has emerged as the standard for training reasoning capabilities in large language models through reinforcement learning. By estimating advantages using group-mean rewards rather than a learned…

Artificial Intelligence · Computer Science 2026-03-06 Anisha Garg , Claire Zhang , Nishit Neema , David Bick , Ganesh Venkatesh , Joel Hestness

Inference-time scaling through multiple sample generation in combination with Process- or Outcome-Reward Model (PRM or ORM) re-ranking has proven effective for text-based reasoning in large language models. This paper investigates whether…

Computation and Language · Computer Science 2025-10-20 Minghan Wang , Thuy-Trang Vu , Ehsan Shareghi , Gholamreza Haffari

Although Large Reasoning Models (LRMs) have progressed in solving complex problems, their chain-of-thought (CoT) reasoning often contains harmful content that can persist even when the final responses appear safe. We show that this issue…

Artificial Intelligence · Computer Science 2026-03-03 Yichi Zhang , Yue Ding , Jingwen Yang , Tianwei Luo , Dongbai Li , Ranjie Duan , Qiang Liu , Hang Su , Yinpeng Dong , Jun Zhu

Implicit in-context learning (ICL) has newly emerged as a promising paradigm that simulates ICL behaviors in the representation space of Large Language Models (LLMs), aiming to attain few-shot performance at zero-shot cost. However,…

Computation and Language · Computer Science 2025-09-30 Jiaqian Li , Yanshu Li , Ligong Han , Ruixiang Tang , Wenya Wang

Chain-of-Thought (CoT) reasoning successfully enhances the reasoning capabilities of Large Language Models (LLMs), yet it incurs substantial computational overhead for inference. Existing CoT compression methods often suffer from a critical…

Machine Learning · Computer Science 2026-05-26 Yuntian Tang , Bohan Jia , Wenxuan Huang , Lianyue Zhang , Jiao Xie , Wenxi Li , Wei Li , Jie Hu , Xinghao Chen Rongrong Ji , Shaohui Lin

Inverse reinforcement learning (IRL) for linear systems seeks a cost function whose optimal controller reproduces an expert policy from data. Existing data-driven methods for discrete-time linear systems are largely built on iterative…

Systems and Control · Electrical Eng. & Systems 2026-05-12 Duc Cuong Nguyen , Phuong Nam Dao

We consider the problem of learning the best possible policy from a fixed dataset, known as offline Reinforcement Learning (RL). A common taxonomy of existing offline RL works is policy regularization, which typically constrains the learned…

Machine Learning · Computer Science 2023-08-16 Yuhang Ran , Yi-Chen Li , Fuxiang Zhang , Zongzhang Zhang , Yang Yu

Current LLM post-training methods optimize complete reasoning trajectories through Supervised Fine-Tuning (SFT) followed by outcome-based Reinforcement Learning (RL). While effective, a closer examination reveals a fundamental gap: this…

Artificial Intelligence · Computer Science 2026-05-29 Shaojie Wang , Liang Zhang