Author

Skylar Zhai

Observed name: Zhai, Skylar

Institutions: University of Southern California, University of Minnesota, Mohamed bin Zayed University of Artificial Intelligence, McGill University, University of Pennsylvania

Observed institutions: University of Southern California, University of Minnesota, Mohamed bin Zayed University of Artificial Intelligence, McGill University, University of Pennsylvania

OpenAlex ID: A5134207559

The Blind Spot of Agent Safety: How Benign User Instructions Expose Critical Vulnerabilities in Computer-Use Agents

2026 · Ding, Xuwei, Zhai, Skylar, et al. · arXiv (Cornell University) · 2

Abstain-R1: Calibrated Abstention and Post-Refusal Clarification via Verifiable RL

2026 · Zhai, Skylar, Liang, Jingcheng, et al. · arXiv (Cornell University) · 1

Each Reward to Its Relevant Tokens: Reward-Specific Token-Level Credit Assignment for Multi-Reward RL

2026 · Zhai, Skylar, Wang, Haonan, et al. · arXiv (Cornell University)

Make Sparse Rewards Count: Density-Aware Reward Aggregation for Multi-Reward RL

2026 · Zheng, Tong, Zhai, Skylar, et al. · arXiv (Cornell University)

ThinkPrior: Zero-Rollout Difficulty Priors for Cold-Start Prompt Selection in RLVR

2026 · Sha, Tommy, Zhai, Skylar, et al. · arXiv (Cornell University)

Works count: 5

Citation count: 3

H-index: 1

I10-index: