10 papers
Pause and Think: A Dataset and Benchmark for Video-Grounded Assistive Action Suggestion
Shivam Singh, Saptarshi Majumder, Pratik Prabhanjan Brahma +2
Recent Vision-Language Models (VLMs) struggle with grounded reasoning, temporal consistency, and context aware planning in videos. We introduce pause-and-think-T, a reasoning-centr…
PR2: Predictive Routing Replay for MoE-Based LLM Reinforcement Learning
Daize Dong, Junlin Chen, Haolong Jia +9
Mixture of Experts (MoE) Large Language Models (LLMs) achieve strong performance at scale. However, reinforcement learning (RL) on MoE-based LLMs often suffers from training instab…
Beyond the Target: From Imitation to Collaboration in Speculative Decoding
Jinze Li, Yixing Xu, Guanchen Li +7
Speculative decoding (SPD) accelerates large language model (LLM) inference by letting a smaller draft model propose multiple future tokens that are verified in parallel by a large…
AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents
Sharareh Younesian, Wenwen Ouyang, Sina Rafati +11
GPU kernel optimization is increasingly critical for efficient deep learning systems, but writing high-performance kernels still requires substantial low-level expertise. Recent AI…
Theory-optimal Quantization Based on Flatness
Xiusheng Huang, Zhe Li, Xuanwu Yin +5
Post-training quantization has emerged as a widely adopted technique for compressing and accelerating the inference of Large Language Models (LLMs). The primary challenges in LLMs…
DC-DiT: Adaptive Compute and Elastic Inference for Visual Generation via Dynamic Chunking
Akash Haridas, Utkarsh Saxena, Parsa Ashrafi Fashi +3
Diffusion Transformers rely on static patchify tokenization, assigning the same token budget to smooth backgrounds, detailed object regions, noisy early timesteps, and late-stage r…