Related papers: Self-Trained Verification for Training- and Test-T…
Large vision-language models (VLMs) increasingly adopt post-training techniques such as long chain-of-thought (CoT) supervised fine-tuning (SFT) and reinforcement learning (RL) to elicit sophisticated reasoning. While these methods exhibit…
Language model (LM) post-training relies on two stages of human supervision: task demonstrations for supervised finetuning (SFT), followed by preference comparisons for reinforcement learning from human feedback (RLHF). As LMs become more…
Can a model learn to escape its own learning plateau? Reinforcement learning methods for finetuning large reasoning models stall on datasets with low initial success rates, and thus little training signal. We investigate a fundamental…
Reinforcement Learning with Verifiable Rewards (RLVR) has recently demonstrated notable success in enhancing the reasoning performance of large language models (LLMs), particularly on mathematics and programming tasks. Similar to how…
This paper investigates the efficacy of a regularized multi-task learning (MTL) framework based on SVM (M-SVM) to answer whether MTL always provides reliable results and how MTL outperforms independent learning. We first find that M-SVM is…
Reinforcement Learning with Verifiable Rewards (RLVR) improves final-answer accuracy on reasoning tasks, but it does not reliably improve reasoning quality. Because outcome rewards only assess final answers, they also reward spurious…
Verification-guided self-improvement has recently emerged as a promising approach to improving the accuracy of large language model (LLM) outputs. However, existing approaches face a trade-off between inference efficiency and accuracy:…
While reinforcement learning (RL) demonstrated remarkable success in enhancing the reasoning capabilities of language models, the training dynamics of RL in LLMs remain unclear. In this work, we provide an explanation of the RL training…
Reinforcement Learning with Verifiable Rewards (RLVR) has achieved great success in developing Large Language Models (LLMs) with chain-of-thought rollouts for many tasks such as math and coding. Nevertheless, RLVR struggles with sample…
We show that reinforcement learning with verifiable rewards (RLVR) can elicit strong mathematical reasoning in certain language models even with spurious rewards that have little, no, or even negative correlation with the correct answer.…
Scaling test-time compute has emerged as a powerful mechanism for enhancing Large Language Model (LLM) performance. However, standard post-training paradigms, Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), optimize the…
Recent advances in vision-language models (VLMs) have improved Chest X-ray (CXR) interpretation in multiple aspects. However, many medical VLMs rely solely on supervised fine-tuning (SFT), which optimizes next-token prediction without…
While state-of-the-art vision-language models (VLMs) have demonstrated remarkable capabilities in complex visual-text tasks, their success heavily relies on massive model scaling, limiting their practical deployment. Small-scale VLMs offer…
Large language models have achieved significant reasoning improvements through reinforcement learning with verifiable rewards (RLVR). Yet as model capabilities grow, constructing high-quality reward signals becomes increasingly difficult,…
The reasoning abilities of large language models (LLMs) have improved with chain-of-thought (CoT) prompting, allowing models to solve complex tasks stepwise. However, training CoT capabilities requires detailed reasoning data, which is…
Test time scaling is currently one of the most active research areas that shows promise after training time scaling has reached its limits. Deep-thinking (DT) models are a class of recurrent models that can perform easy-to-hard…
Safe deployment of large language models (LLMs) may benefit from a reliable method for assessing their generated content to determine when to abstain or to selectively generate. While likelihood-based metrics such as perplexity are widely…
The success of RL for LLM post-training stems from an unreasonably uninformative source: a single bit of information per rollout as binary reward or preference label. At the other extreme, distillation offers dense supervision but requires…
Reinforcement Learning with Verifiable Rewards (RLVR) improves LLM reasoning, yet growing evidence indicates an exploration ceiling: it often reweights existing solution traces rather than discovering new strategies, limiting gains under…
Training on model-generated synthetic data is a promising approach for finetuning LLMs, but it remains unclear when it helps or hurts. In this paper, we investigate this question for math reasoning via an empirical study, followed by…