Taehyun Cho
Scholar

Taehyun Cho

Google Scholar ID: kVi85ZgAAAAJ
Seoul National University
Reinforcement Learning
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
51
 
H-index
5
 
i10-index
2
 
Publications
14
 
Co-authors
0
 
Contact
Emailtalium@cml.snu.ac.krCVOpen ↗GitHubOpen ↗LinkedInOpen ↗
Publications
6 items
A Regret Minimization Framework on Preference Learning in Large Language Models
2026
Cited
0
Probabilistic Smoothing with Ratio-Monotone Transforms for Global Optimization
2026
Cited
0
Why Latent Actions Fail, and How to Prevent It
2026
Cited
0
MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction
2026
Cited
0
Policy-labeled Preference Learning: Is Preference Enough for RLHF?
2025
Cited
0
Bellman Unbiasedness: Tractable and Provably Efficient Distributional Reinforcement Learning with General Value Function Approximation
2024
Cited
0
Resume
Academic Achievements
  • - Publications:
  • - 'Policy-labeled Preference Learning: Is Preference Enough for RLHF?' (ICML 2025)
  • - 'Bellman Unbiasedness: Toward Provably Efficient Distributional Reinforcement Learning with General Value Function Approximation' (ICML 2025)
  • - 'Spectral-Risk Safe Reinforcement Learning with Convergence Guarantees' (NeurIPS 2024)
  • - 'Pitfall of Optimism: Distributional Reinforcement Learning by Randomizing Risk Criterion' (NeurIPS 2023)
  • - 'SPQR: Controlling Q-ensemble Independence with Spiked Random Model for Reinforcement Learning' (NeurIPS 2023)
  • - 'On the Convergence of Continual Learning with Adaptive Methods' (UAI 2023)
  • - 'Adaptive Methods for Nonconvex Continual Learning' (NeurIPS 2022)
  • - 'Perturbed Quantile Regression for Distributional Reinforcement Learning' (NeurIPS 2022)
  • - 'Chebyshev polynomial codes: Task entanglement-based coding for distributed matrix multiplication' (ICML 2021)
  • - 'Optimized shallow neural networks for sum-rate maximization in energy harvesting downlink multiuser NOMA systems' (IEEE Journal on Selected Areas in Communications)
  • - 'An Efficient Neural Network Architecture for Rate Maximization in Energy Harvesting Downlink Channels' (2020 IEEE International Symposium on Information Theory (ISIT))
Research Experience
  • - Research Projects:
  • - Preprints: 'An Axiomatization of Process Score Model', etc.
  • - Work In Progress: 'Off-policy Direct Preference Optimization with Monotonic Improvement Guarantee', etc.
  • - Work In Progress: 'Policy Optimization with Process Regret Model', etc.
  • - Position: Ph.D. Candidate
Education
  • - Degree: Ph.D. Candidate
  • - University: Seoul National University
  • - Advisor: Jungwoo Lee
  • - Year: Final year
  • - Department: Electrical and Computer Engineering
  • - Bachelor's Degree: Mathematics
  • - University: Korea University
Background
  • - Research Interests: Sequential decision-making under uncertainty, particularly in the context of human feedback
  • - Specialization: Distributional reinforcement learning (distRL), reinforcement learning from human feedback (RLHF), and regret analysis
  • - Goal: To develop mathematical models and optimize for human-in-the-loop systems, uncovering both theoretical insights and practical algorithms for robust decision-making
  • - Current Interests: Reasoning LLM agents and regret-based decision theory
Miscellany
  • - Personal Interests: Actively looking for postdoctoral opportunities in theoretical foundations of reinforcement learning or reasoning LLM research