Yike Guo

From bits to atoms: A survey of cross-layer safety in Embodied AI

Embodied Artificial Intelligence (AI) introduces cyber–physical risks because digital failures and attacks can cause physical harm. This survey organizes embodied AI safety through a four-layer reference architecture — comprising the physical interaction, system and middleware, algorithm and decision, and cloud–edge an...

CoBrush: A Hierarchical Planning Framework for Human-Robot Co-Painting

Embodied co-painting requires a robot to repeatedly update a shared physical canvas while human intent evolves over interaction. Existing reference-driven painters or reactive assistants are typically optimized for single-shot rendering or sketch completion, limiting their ability to sustain coherent multi-round collab...

A Theory of Reliable Self-Evolution for Agent Harnesses

In harness self-evolution, agents modify their own prompts, code, tools, and orchestration while keeping the underlying language model fixed. Recent work has shown that agents can improve themselves in response to task failures and achieve substantial performance gains. However, gains on failed tasks do not automatical...

OMP-MoE: Efficient Expert Pruning for Mixture-of-Experts LLMs via Orthogonal Matching Pursuit

OMP-MoE, a novel training-free compression framework for reducing expert redundancy in MoE-based LLMs by reformulate the pruning problem as a sparse signal reconstruction task solved through Orthogonal Matching Pursuit.

Can Coding Agents Solve Repository-Level Issues with Rendered Code? An Exploratory Study of Visual Representations

Visual modality has recently been explored as a way to compress textual tokens, including rendering code as images for static code understanding. We study whether this representation can serve as operational context for agentic coding, where an agent must navigate repositories, edit source files, and verify executable...

PASTABench: Proactive Assessment of Sequential Trajectories for Agent Safety

As Large Language Models (LLMs) evolve into autonomous agents that alter real-world states, ensuring operational safety across multi-step workflows has become a critical challenge. While recent work has moved beyond single-turn evaluation toward multi-turn paradigms, key limitations persist: step-level methods treat ac...

MD-VAE: Concepts embedded variational autoencoder with multiple decoders

LexAgentHallu: A Hierarchical Benchmark for Profiling Hallucinations in Legal Agents

As large language models are increasingly deployed as tool-augmented legal agents, they introduce agentic hallucinations where tool-call and reasoning errors cascade into fabricated holdings and miscited authority. However, existing legal benchmarks evaluate only single-turn QA with outcome-level metrics, while agentic...

LUNA: Learning Universal 3D Human Animation Beyond Skinning

LUNA is the first end-to-end 3D animatable model that supports implicit 2D driving and introduces hybrid supervision that distills soft structural priors from an LBS teacher and a loss that supports training on both limited fitted data and large in-the-wild unlabeled videos.

Life Operators: a self-evolving framework for multiscale life modelling

Life Operators is proposed: task-bounded mappings that define three scientific roles: Perception operators infer task-relevant biological states from multimodal observations, Evolution operators propagate these states under natural or intervention-conditioned dynamics, and Generation operators map them to measurable si...

Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence

This analysis provides a structured account of current approaches to scaling LRMs beyond human supervision and the open problems involved in developing self-sustaining learning systems toward superintelligence.

Zero2Skill: Bootstrapping Robot Skills through Autonomous Data Collection, Training, and Deployment

Zero2Skill is presented, a human-robot symbiotic agentic system in which corrections are retained and reused across rounds, and policies fine-tuned on Zero2Skill data match teleoperation-trained policy success at a fraction of collection human cost.