Bo Liu

This paper systematically revisits and expands the threat model for backdoor attacks on semantic segmentation, proposing a unified framework (BADSEG) that demonstrates severe, previously overlooked vulnerabilities in current and emerging segmentation models.

The paper proposes a comprehensive framework for LLM-based agent unlearning, enabling agents to selectively forget specific knowledge (states, trajectories, or environments) while maintaining performance and resisting knowledge inference by adversaries.

The paper proposes Cluster Segregation Concealment (CSC), a novel defense that identifies and neutralizes backdoor triggers by relabeling poisoned samples to a virtual class, achieving near-zero attack success rates with minimal accuracy loss.

The paper introduces the Agentic Bio-Capabilities Benchmark (ABC-Bench) to measure biosecurity-relevant capabilities of large language models.

This paper proposes a new method for cooperative multi-objective multi-agent reinforcement learning that learns agent-specific preferences to improve team performance and trade-off coordination.

This paper presents a human-in-the-loop segmentation framework, Hi-Seg, that improves medical segmentation accuracy while reducing annotation time for medical annotators.

This paper studies how scientists identify relevant prior research and develops a new dataset, ScholarCatalyst, to aid in this process.

DMAD recasts distribution matching as classification and learns log-density ratios directly to improve few-step student training.

The paper introduces FrontierPhysics, an end-to-end benchmark for evaluating AI agents in physics research.