Related papers: Group Relative Policy Optimization for Text-to-Spe…
Reinforcement learning (RL) has become popular in enhancing the reasoning capabilities of large language models (LLMs), with Group Relative Policy Optimization (GRPO) emerging as a widely used algorithm in recent systems. Despite GRPO's…
Existing methods for extracting reward signals in Reinforcement Learning typically rely on labeled data and dedicated training splits, a setup that contrasts with how humans learn directly from their environment. In this work, we propose…
Current Large Language Models (LLMs) often undergo supervised fine-tuning (SFT) to acquire tool use capabilities. However, SFT struggles to generalize to unfamiliar or complex tool use scenarios. Recent advancements in reinforcement…
Effective information seeking in multi-turn medical dialogues is critical for accurate diagnosis, especially when dealing with incomplete information. Aligning Large Language Models (LLMs) for these interactive scenarios is challenging due…
Large language models are increasingly trained via reinforcement learning for personalized recommendation tasks, but standard methods like GRPO rely on sparse, sequence-level rewards. These obscure which tokens actually contribute to…
Enhancing the conformity of large language models (LLMs) to human preferences remains an ongoing research challenge. Recently, offline approaches such as Direct Preference Optimization (DPO) have gained prominence as attractive options due…
Reinforcement learning from verifiable rewards has emerged as a powerful technique for enhancing the complex reasoning abilities of Large Language Models (LLMs). However, these methods are fundamentally constrained by the ''learning cliff''…
Aligning large language models (LLMs) to human values has become increasingly important as it enables sophisticated steering of LLMs. However, it requires significant human demonstrations and feedback or distillation from proprietary LLMs…
Large Language Models (LLMs) have made remarkable progress in enhancing step-by-step reasoning through reinforcement learning. However, the Group Relative Policy Optimization (GRPO) algorithm, which relies on sparse reward rules, often…
Group Relative Policy Optimization (GRPO), recently introduced by DeepSeek, is a critic-free reinforcement learning algorithm for fine-tuning large language models. GRPO replaces the value function in Proximal Policy Optimization (PPO) with…
Training Large Language Models (LLMs) with Group Relative Policy Optimization (GRPO) encounters a significant challenge: models often fail to produce accurate responses, particularly in small-scale architectures. This limitation not only…
Traditional on-policy Reinforcement Learning with Verifiable Rewards (RLVR) frameworks suffer from experience waste and reward homogeneity, which directly hinders learning efficiency on difficult samples during large language models…
Bus holding control is a widely-adopted strategy for maintaining stability and improving the operational efficiency of bus systems. Traditional model-based methods often face challenges with the low accuracy of bus state prediction and…
Large Language Models (LLMs) have demonstrated impressive capabilities in language processing, yet they often struggle with tasks requiring genuine visual spatial reasoning. In this paper, we introduce a novel two-stage training framework…
Signal Temporal Logic (STL) is a powerful formal language for specifying real-time specifications of Cyber-Physical Systems (CPS). Transforming specifications written in natural language into STL formulas automatically has attracted…
Aligning Large Language Models (LLMs) with human preferences typically relies on external supervision, which faces critical limitations: human annotations are scarce and subjective, reward models are vulnerable to reward hacking, and…
Text-to-speech (TTS) models have been widely adopted to enhance automatic speech recognition (ASR) systems using text-only corpora, thereby reducing the cost of labeling real speech data. Existing research primarily utilizes additional text…
Simultaneous speech translation (SST) generates translations while receiving partial speech input. Recent advances show that large language models (LLMs) can substantially improve SST quality, but at the cost of high computational overhead.…
Speech-to-text alignment is a critical component of neural textto-speech (TTS) models. Autoregressive TTS models typically use an attention mechanism to learn these alignments on-line. However, these alignments tend to be brittle and often…
Advances in large language models (LLMs) are driving a shift toward using reinforcement learning (RL) to train agents from iterative, multi-turn interactions across tasks. However, multi-turn RL remains challenging as rewards are often…