243 citations · 268 across the 14 of their papers we have counts for
15 papers
Router Drift Control: Mitigating Off-Policy Degradation in Mixture-of-Experts Reinforcement Learning
Liangyao Jin, Zhixin Wang, Liming Liu +8
Reinforcement learning (RL) can become off-policy due to training-inference mismatch, mini-batch updates, and asynchronous training. The resulting mismatch is especially harmful in…
FalconGEMM: Surpassing Hardware Peaks with Lower-Complexity Matrix Multiplication
Honglin Zhu, Jiaping Cao, Jiang Shao +10
Peak breaking Matrix Multiplication is a promising technique to improve the performance of DL, especially in LLM training and inference. We present FalconGEMM, a cross-platform fra…
PlexRL: Cluster-Level Orchestration of Serviceized LLM Execution for RLVR
Yiqi Zhang, Fangzheng Jiao, Tian Tang +13
Reinforcement learning with verifiable rewards (RLVR) has recently unlocked strong reasoning capabilities in large language models (LLMs), triggering rapid exploration of new algor…
daVinci-LLM:Towards the Science of Pretraining
Yiwei Qin, Yixiu Liu, Tiantian Mi +12
The foundational pretraining phase determines a model's capability ceiling, as post-training struggles to overcome capability foundations established during pretraining, yet it rem…
Data Darwinism Part I: Unlocking the Value of Scientific Data for Pre-training
Yiwei Qin, Zhen Huang, Tiantian Mi +5
Data quality determines foundation model performance, yet systematic processing frameworks are lacking. We introduce Data Darwinism, a ten-level taxonomy (L0-L9) that conceptualize…
Nex-N1: Agentic Models Trained via a Unified Ecosystem for Large-Scale Environment Construction
AGI Team, Yuxuan Cai, Lu Chen +62
The evolution of Large Language Models (LLMs) from passive responders to autonomous agents necessitates a fundamental shift in learning paradigms -- from static imitation to incent…
ReSpec: Towards Optimizing Speculative Decoding in Reinforcement Learning Systems
Qiaoling Chen, Zijun Liu, Peng Sun +6
Adapting large language models (LLMs) via reinforcement learning (RL) is often bottlenecked by the generation stage, which can consume over 75\% of the training time. Speculative d…
Expert-as-a-Service: Towards Efficient, Scalable, and Robust Large-scale MoE Serving
Ziming Liu, Boyu Tian, Guoteng Wang +15
Mixture-of-Experts (MoE) models challenge serving infrastructures with dynamic, sparse expert utilization, causing instability on conventional systems designed for dense architectu…
DistFlow: A Fully Distributed RL Framework for Scalable and Efficient LLM Post-Training
Zhixin Wang, Jiaming Xu, Tianyi Zhou +10
Effectively scaling Reinforcement Learning (RL) is crucial for enhancing the reasoning and alignment of Large Language Models. The massive data and complex execution flows inherent…
WebLLM: A High-Performance In-Browser LLM Inference Engine
Charlie F. Ruan, Yucheng Qin, Akaash R. Parthasarathy +11
Advancements in large language models (LLMs) have unlocked remarkable capabilities. While deploying these models typically requires server-grade GPUs and cloud-based inference, the…
Productively Deploying Emerging Models on Emerging Platforms: A Top-Down Approach for Testing and Debugging
Siyuan Feng, Jiawei Liu, Ruihang Lai +4
While existing machine learning (ML) frameworks focus on established platforms, like running CUDA on server-grade GPUs, there have been growing demands to enable emerging AI applic…
Relax: Composable Abstractions for End-to-End Dynamic Machine Learning
Ruihang Lai, Junru Shao, Siyuan Feng +16
Dynamic shape computations have become critical in modern machine learning workloads, especially in emerging large language models. The success of these models has driven the deman…
TensorIR: An Abstraction for Automatic Tensorized Program Optimization
Siyuan Feng, Bohan Hou, Hongyi Jin +8
Deploying deep learning models on various devices has become an important topic. The wave of hardware specialization brings a diverse set of acceleration primitives for multi-dimen…
Tensor Program Optimization with Probabilistic Programs
Junru Shao, Xiyou Zhou, Siyuan Feng +7
Automatic optimization for tensor programs becomes increasingly important as we deploy deep learning in various environments, and efficient optimization relies on a rich search spa…
CityFlow: A Multi-Agent Reinforcement Learning Environment for Large Scale City Traffic Scenario
Huichu Zhang, Siyuan Feng, Chang Liu +7
Traffic signal control is an emerging application scenario for reinforcement learning. Besides being as an important problem that affects people's daily life in commuting, traffic…