5 papers
Mergeable Model-Side Aggregation States for Long-Context Language Models
Dachuan Song, Junyu Yin, Zechen Hu +1
The paper proposes a model-side aggregation interface that uses compact HyperLogLog sketches to maintain set-based aggregation states alongside frozen language models, enabling acc…
Smooth Scaling Laws Hide Stepwise Token Learning
Pingjie Wang, Zechen Hu, Peiru Yang +2
Language model loss follows remarkably regular scaling laws over model and data size, yet it remains unclear why the aggregate loss should exhibit a power-law form. Existing explan…
Efficient Multi-turn RL for GUI Agents via Decoupled Training and Adaptive Data Curation
Pengxiang Li, Zechen Hu, Zirui Shang +15
Vision-language model (VLM) based GUI agents show promise for automating complex desktop and mobile tasks, but face significant challenges in applying reinforcement learning (RL):…
Robust Online Calibration for UWB-Aided Visual-Inertial Navigation with Bias Correction
Yizhi Zhou, Jie Xu, Jiawei Xia +3
This paper presents a novel robust online calibration framework for Ultra-Wideband (UWB) anchors in UWB-aided Visual-Inertial Navigation Systems (VINS). Accurate anchor positioning…
CARoL: Context-aware Adaptation for Robot Learning
Zechen Hu, Tong Xu, Xuesu Xiao +1
Using Reinforcement Learning (RL) to learn new robotic tasks from scratch is often inefficient. Leveraging prior knowledge has the potential to significantly enhance learning effic…