1 citations · 1 across the 12 of their papers we have counts for
12 papers
LEAP: Lean Environment-Feedback via Adaptive Pruning for Code RL in GPU Kernel Generation
Tankun Li, Zhi Chen, Yaohua Tang
Post-training large language models (LLMs) via reinforcement learning (RL) has significantly advanced code generation capabilities. To bypass the heavy memory footprint of critic n…
TCPO: Turn-Level Credit Policy Optimization
Sicong Liao, Zhi Chen, Yaohua Tang
Verifier-guided reinforcement learning has become a powerful paradigm for improving LLM reasoning. In multi-turn settings, models receive a verifier score after each turn and itera…
Beyond the Best Teacher: Expanding and Compressing the Reasoning Solution Manifold
Songshuo Lu, Zhi Chen, Yaohua Tang
A single reinforcement-learning run can produce a strong reasoner yet an incomplete teacher: it often amplifies only a subset of the valid solution modes. We argue that reinforceme…
Latent Reasoning with Normalizing Flows
Guancheng Tu, Xiangjun Fu, Suhao Yu +5
Large language models often improve reasoning by generating explicit chain-of-thought (CoT), demonstrating the importance of intermediate computation. However, textual CoT forces t…
MusaCoder: Native GPU Kernel Generation with Full-Stack Training on Moore Threads GPU
Kun Cheng, Songshuo Lu, Sicong Liao +7
Native GPU kernel generation turns high-level tensor programs into executable, efficient low-level code. Existing Large Language Models (LLMs) struggle with this task, while execut…
LiteGUI: Lightweight GUI Agents via Multi-Solution Guided Distillation and Dual-Level Reinforcement Learning
Yubin Wu, Zicheng Cai, Liping Ning +4
We present LiteGUI, a new framework for building lightweight GUI agents. GUI interaction poses unique challenges due to the long-horizon nature of complex tasks and the existence o…
UniSem: Generalizable Semantic 3D Reconstruction from Sparse Unposed Images
Guibiao Liao, Qian Ren, Kaimin Liao +4
Semantic-aware 3D reconstruction from sparse, unposed images remains challenging for feed-forward 3D Gaussian Splatting (3DGS). Existing methods often predict an over-complete set…
URPO: A Unified Reward & Policy Optimization Framework for Large Language Models
Songshuo Lu, Hua Wang, Zhi Chen +1
Large-scale alignment pipelines typically pair a policy model with a separately trained reward model whose parameters remain frozen during reinforcement learning (RL). This separat…
StableGS: A Floater-Free Framework for 3D Gaussian Splatting
Luchao Wang, Qian Ren, Kaimin Liao +3
3D Gaussian Splatting (3DGS) reconstructions are plagued by stubborn ``floater" artifacts that degrade their geometric and visual fidelity. We are the first to reveal the root caus…
Round Attention: A Novel Round-Level Attention Mechanism to Accelerate LLM Inference
Yaohua Tang, Zhicheng Hu, Kun Cheng +4
The increasing context window size in large language models (LLMs) has improved their ability to handle complex, long-text tasks. However, as the conversation rounds continue, it i…
TurboRAG: Accelerating Retrieval-Augmented Generation with Precomputed KV Caches for Chunked Text
Songshuo Lu, Hua Wang, Yutian Rong +2
Current Retrieval-Augmented Generation (RAG) systems concatenate and process numerous retrieved document chunks for prefill which requires a large volume of computation, therefore…
A Full-duplex Speech Dialogue Scheme Based On Large Language Models
Peng Wang, Songshuo Lu, Yaohua Tang +3
We present a generative dialogue system capable of operating in a full-duplex manner, allowing for seamless interaction. It is based on a large language model (LLM) carefully align…