publications

2026

1. arXiv 2026

   MonitorBench: A Comprehensive Benchmark for Chain-of-Thought Monitorability in Large Language Models

   Han Wang^*, Yifan Sun^*, Brian Ko^*, Mann Talati, Jiawen Gong, Zimeng Li, Naicheng Yu, Xucheng Yu, Wei Shen, Vedant Jolly, and Huan Zhang

   2026

   PDF
2. ICLR 26

   On The Fragility of Benchmark Contamination Detection in Reasoning Models

   Han Wang^*, Haoyu Li^*, Brian Ko^*, and Huan Zhang

   2026

   PDF

2025

1. Pre-Print

   A Unified Framework for Comparing Distribution Matching Methods Across Trustworthy Machine Learning Tasks

   Brian Ko, Ziyu Gong, Jim Lim, and David Inouye

   2025

   PDF

2023

1. IEEE RO-MAN Oral

   Backward Curriculum Reinforcement Learning

   Brian Ko

   In 2023 32nd IEEE International Conference on Robot and Human Interactive Communication (RO-MAN), 2023

   PDF Code
2. Pre-Print

   Unmasking the Author: Exploiting Code Language Models and Contrastive Learning in Binary Code Authorship Attribution

   Kyung Min Ko, Nan Jiang, and Lin Tan

   2023

   PDF Code