LARK Lab@HKUST (GZ)

non-profit

https://lark-lab-hkustgz.github.io/

AI & ML interests

Large Language Models

Recent Activity

terr1ble authored a paper 1 day ago

SafeBuild-Bench: A Temporal-Robust Construction Safety Benchmark with Graph-Enhanced Data Mining

terr1ble authored a paper 1 day ago

Trajectory-Level Speculation with Semantic Verification for Tool-Using Agents

sapphirex submitted a paper 2 days ago

Mosaic: Multi-Session Agentic Reinforcement Learning

View all activity

Papers

Mosaic: Multi-Session Agentic Reinforcement Learning

From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning

View all Papers

LARK-Lab 's papers 5

Submitted by

yilong xu

Mosaic: Multi-Session Agentic Reinforcement Learning

LARK-Lab

LARK Lab@HKUST (GZ)

Submitted by

Chen chao

From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning

LARK-Lab

LARK Lab@HKUST (GZ)

Submitted by

Zhou

Attention Amnesia in Hybrid LLMs: When CoT Fine-Tuning Breaks Long-Range Recall, and How to Fix It

LARK-Lab

LARK Lab@HKUST (GZ)

Submitted by

shawnxzhu

EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL

LARK-Lab

LARK Lab@HKUST (GZ)

Submitted by

Zhou

Efficient RLVR Training via Weighted Mutual Information Data Selection

LARK-Lab

LARK Lab@HKUST (GZ)

2