Xin-Qiang Cai
Scholar

Xin-Qiang Cai

Google Scholar ID: rtMUMooAAAAJ
RIKEN Center for Advanced Intelligence Project
Machine LearningReinforcement LearningImitation Learning
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
152
 
H-index
8
 
i10-index
7
 
Publications
20
 
Co-authors
0
 
Contact
GitHubOpen ↗
Publications
8 items
Estimating and Orthogonalizing Unknown Pre-training Gradients for Continual Fine-tuning of Large Language Models
2026
Cited
0
VI-CuRL: Stabilizing Verifier-Independent RL Reasoning via Confidence-Guided Variance Reduction
2026
Cited
0
Positive-Unlabeled Reinforcement Learning Distillation for On-Premise Small Models
2026
Cited
0
Rethinking Toxicity Evaluation in Large Language Models: A Multi-Label Perspective
2025
Cited
0
Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers
2025
Cited
0
PIG-Nav: Key Insights for Pretrained Image Goal Navigation Models
2025
Cited
0
UC-MOA: Utility-Conditioned Multi-Objective Alignment for Distributional Pareto-Optimality
2025
Cited
0
Offline Reinforcement Learning with Domain-Unlabeled Data
2024
Cited
0
Resume
Background
  • Research Interests: Reward Modeling, Reinforcement Learning & Imitation Learning, Learning with Weak Supervision. Biography: Currently a postdoctoral researcher in the Imperfect Information Learning Team at RIKEN Center for Advanced Intelligence Project (AIP), led by Professor Masashi Sugiyama.
Miscellany
  • Recent activities include attending workshops between the University of Melbourne and RIKEN-AIP in Melbourne (July 3–4, 2025) and the University of Sydney and RIKEN-AIP in Sydney (July 7–8, 2025).