14 papers
LoGRA: Scaling LLM Reinforcement Learning with Low-Rank Gradient Sketches
Shaokun Zhang, Yifan Zhang, Jian Hu +5
Reinforcement learning has greatly advanced the capabilities of large language models, but its memory demands remain a barrier to broader adoption. We introduce LoGRA, an approach…
PhyAI: Real-Time Physical AI at the Edge, Scalable Rollouts in the Cloud
Chenghua Wang, Daliang Xu, Dongqi Cai +23
Physical AI policies require inference throughout their lifecycle, including model evaluation, cloud reinforcement learning rollout, edge GPU serving, and onboard deployment. Altho…
Molt: A Scalable PyTorch-Native Training Framework for Agentic Reinforcement Learning
Jian Hu, Huiying Li, Hao Zhang +8
Agentic reinforcement learning requires rapid experimentation with agents and learning algorithms, yet large policies and long, multimodal trajectories demand substantial distribut…
AutoLab: Can Frontier Models Solve Long-Horizon Auto Research and Engineering Tasks?
Zhangchen Xu, Junda Chen, Yue Huang +16
Scientific and engineering progress is fundamentally a long-horizon iterative process: proposing changes, running experiments, measuring outcomes, and continuously refining artifac…
Polar: Agentic RL on Any Harness at Scale
Binfeng Xu, Hao Zhang, Shaokun Zhang +9
Reinforcement learning for language agents increasingly depends on custom harnesses that manage long-running context, multi-turn tool use and multi-agent orchestration. However, po…
FastKernels: Benchmarking GPU Kernel Generation in Production
Gabriele Oliaro, Yichao Fu, Jaeseong Lee +8
LLM-based agents for GPU kernel generation are advancing rapidly, but the benchmarks they optimize against evaluate kernels in isolation, with synthetic inputs and weak baselines,…
RW-TTT: Batched Serving for Request-Owned Test-Time Training State
Jian Yang, Zhizhuo Kou, Yao Tian +4
Test-time training (TTT) adapts an LLM during generation by reading and updating request-owned state, such as fast weights, low-rank deltas, or streaming learner state. This breaks…
The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence
Aili Chen, Aonian Li, Baichuan Zhou +215
We introduce the MiniMax-M2 series, a family of Mixture-of-Experts language models built around the principle that mini activations can unleash maximum real-world intelligence. The…
ProRL Agent: Rollout-as-a-Service for RL Training of Multi-Turn LLM Agents
Hao Zhang, Mingjie Liu, Shaokun Zhang +10
Multi-turn LLM agents are increasingly important for solving complex, interactive tasks, and reinforcement learning (RL) is a key ingredient for improving their long-horizon behavi…
Compute Only Once: UG-Separation for Efficient Large Recommendation Models
Hui Lu, Zheng Chai, Shipeng Bai +15
Driven by scaling laws, recommender systems increasingly rely on larger-scale models to capture complex feature interactions and user behaviors, but this trend also leads to prohib…
Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speed
Yonggan Fu, Lexington Whalen, Zhifan Ye +11
Diffusion language models (dLMs) have emerged as a promising paradigm that enables parallel, non-autoregressive generation, but their learning efficiency lags behind that of autore…
SANA-Video: Efficient Video Generation with Block Linear Diffusion Transformer
Junsong Chen, Yuyang Zhao, Jincheng Yu +17
We introduce SANA-Video, a small diffusion model that can efficiently generate videos up to 720x1280 resolution and minute-length duration. SANA-Video synthesizes high-resolution,…
Fast-dLLM v2: Efficient Block-Diffusion LLM
Chengyue Wu, Hao Zhang, Shuchen Xue +7
Autoregressive (AR) large language models (LLMs) have achieved remarkable performance across a wide range of natural language tasks, yet their inherent sequential decoding limits i…
Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding
Chengyue Wu, Hao Zhang, Shuchen Xue +6
Diffusion-based large language models (Diffusion LLMs) have shown promise for non-autoregressive text generation with parallel decoding capabilities. However, the practical inferen…