Kai Chen
7 papers · Latest:
FrameSkip: Learning from Fewer but More Informative Frames in VLA Training
FrameSkip improves VLA policy training by selecting fewer, more informative frames from robot demonstrations, boosting success rates.
WildClawBench: A Benchmark for Real-World, Long-Horizon Agent Evaluation
WildClawBench introduces a new benchmark for evaluating long-horizon, real-world agents using native runtimes and real tools.
EA-WM: Event-Aware Generative World Model with Structured Kinematic-to-Visual Action Fields
EA-WM is a generative world model that uses structured kinematic-to-visual action fields to improve robot interaction dynamics and geometry in generated videos.
STARRY: Spatial-Temporal Action-Centric World Modeling for Robotic Manipulation
STARRY is a novel world model for robotic manipulation that aligns spatial-temporal prediction with action generation for improved task success.
TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based Exploration
TREX is a multi-agent system that automates the entire LLM fine-tuning lifecycle using a tree-based exploration for efficient strategy planning.
Giant Room-Temperature Third-Order Electrical Transport in a Thin-Film Altermagnet Candidate
This paper demonstrates giant room-temperature third-order electrical transport in RuO2 thin films, an altermagnet candidate, driven by quantum geometry.
CrashSight: A Phase-Aware, Infrastructure-Centric Video Benchmark for Traffic Crash Scene Understanding and Reasoning
CrashSight is a new vision-language benchmark using roadside camera data to evaluate AI models' understanding of traffic crashes.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.