collaborators

6 papers

cs.LG2026

Activation Outliers Matter: Robust Recovery for Quantized Multimodal LLMs

Tanzila Rahman, Mehran Taghian Jazi, Yunke Peng +10

Low-bit quantization offers a promising avenue for reducing the computational and memory demands of Multimodal Large Language Models (MLLMs). Recent hardware support for low-precis…

cs.LG2026

HiFloat4 Format for End-To-End Reinforcement Learning Post-Training of Large Language Models

Hei Yi Mak, Shadan Golestan, Hoang Le +10

We present, to our knowledge, the first end-to-end FP4 RL post-training, in which both the rollout and training policies, including their forward and backward passes, operate at 4-…

cs.CL2026

When Implausible Tokens Get Reinforced: Tail-Aware Credit Calibration for LLM Reinforcement Learning

Xiuyi Lou, Zicheng Xu, Yu-Neng Chuang +4

Reinforcement learning (RL) has achieved remarkable success in enhancing the reasoning capabilities of large language models (LLMs). However, widely used critic-free RL methods rel…

cs.LG2026

Forget Without Compromise: Nexus Sampling for Streaming KV-Cache Eviction Under Fixed Budgets

Duc Duong, Hoang Anh Duy Le, Jianwen Xie +2

Long-context and agentic LLM workloads push the KV cache past any fixed memory budget, forcing the inference stack to permanently evict tokens at every step of a continuous-inferen…

cs.CL2026

Learning at the Right Pace: Adaptive Data Scheduling Improves LLM Reinforcement Learning

Zicheng Xu, Ruixuan Zhang, Yu-Neng Chuang +7

Large Language Models (LLMs) achieve remarkable reasoning capabilities through reinforcement learning (RL) post-training. However, existing RL post-training commonly relies on unif…

cs.LG2026

Scout Before You Attend: Sketch-and-Walk Sparse Attention for Efficient LLM Inference

Hoang Anh Duy Le, Sahil Joshi, Zeyu Yang +2

Self-attention dominates the computational and memory cost of long-context LLM inference across both prefill and decode phases. To address this challenge, we introduce Sketch&Walk…