Rutgers University · 5 records

dateidtitleauthorsvenue↑
2026-10-192610.16071Dream2Discern: Learning from Hard Latent Variants for Runtime Failure Prediction in Multi-Agent SystemsBoxuan Zhang et al.arXiv13
2026-10-012609.39102False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search AgentsMeijia Chen et al.arXiv157
2026-05-132605.09296Micro-Defects Expose Macro-Fakes: Detecting AI-Generated Images via Local Distributional ShiftsBoxuan Zhang et al.arXiv3
2026-05-132605.08504A Single Layer to Explain Them All:Understanding Massive Activations in Large Language ModelsZeru Shi et al.arXiv2
2026-05-122605.08715AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent SystemsBoxuan Zhang et al.arXiv8