Leheng Sheng (盛乐恒)

Ph.D. student, National University of Singapore

CV| Google Scholar| GitHub| Publications

I am a Ph.D. student at the National University of Singapore. Previously, I received an M.S. in Finance (Fintech) from Tsinghua University and a B.E. in Computer Science from Southeast University. My research focuses on LLM reasoning and alignment, representation learning in LLMs, and LLM-based recommender systems / agents.

News

[Feb 2026] Self-Evolving Rubrics and When to Memorize released; When to Memorize accepted by ICML 2026.
[Jan 2026] AlphaSteer accepted by ICLR 2026.
[Sep 2025] On Reasoning Strength Planning accepted by NeurIPS 2025.
[Jan 2025] Language Representations Can Be What Recommenders Need accepted by ICLR 2025 (Oral).
[Mar 2024] Agent4Rec accepted by SIGIR 2024.
[Sep 2023] AdvInfoNCE accepted by NeurIPS 2023.

Selected Publications

* denotes co-first author; † denotes corresponding author. First-author / co-first-author works only — see the full list.

[1] Reinforcing Chain-of-Thought Reasoning with Self-Evolving Rubrics arXiv 2026

Leheng Sheng*, Wenchang Ma*, Ruixin Hong, Xiang Wang, An Zhang, Tat-Seng Chua

[2] When to Memorize and When to Stop: Gated Recurrent Memory for Long-Context Reasoning ICML 2026

Leheng Sheng, Yongtao Zhang, Wenchang Ma, Yaorui Shi, Ting Huang, Xiang Wang, An Zhang, Ke Shen, Tat-Seng Chua

[3] AlphaSteer: Learning Refusal Steering with Principled Null-Space Constraint ICLR 2026

Leheng Sheng*, Changshuo Shen*, Weixiang Zhao, Junfeng Fang, Xiaohao Liu, Zhenkai Liang, Xiang Wang, An Zhang, Tat-Seng Chua

[4] On Reasoning Strength Planning in Large Reasoning Models NeurIPS 2025

Leheng Sheng, An Zhang, Zijian Wu, Weixiang Zhao, Changshuo Shen, Yi Zhang, Xiang Wang, Tat-Seng Chua

[5] Language Representations Can Be What Recommenders Need: Findings and Potentials ICLR 2025 (Oral)

Leheng Sheng, An Zhang†, Yi Zhang, Yuxin Chen, Xiang Wang, Tat-Seng Chua

[6] On Generative Agents in Recommendation SIGIR 2024

An Zhang*, Yuxin Chen*, Leheng Sheng*, Xiang Wang†, Tat-Seng Chua

[7] Empowering Collaborative Filtering with Principled Adversarial Contrastive Loss NeurIPS 2023

An Zhang*, Leheng Sheng*, Zhibo Cai†, Xiang Wang, Tat-Seng Chua

[8] BrainNetFormer: Decoding Brain Cognitive States with Spatial-Temporal Cross Attention ICASSP 2023

Leheng Sheng, Wehan Wang, Zhiyi Shi, Jichao Zhan, Youyong Kong†

Education

National University of Singapore · 2024.8 – present
Ph.D. student, School of Computing

Tsinghua University · 2022.9 – 2024.6
M.S. in Finance (Fintech), School of Economics and Management

Southeast University · 2018.9 – 2022.6
B.E. in Computer Science and Technology, Chien-Shiung Wu College (Honors)

Internships

Research Intern, Pretrain Team · ByteDance Seed · Aug 2025 – present
Topic: General & long-context LLM reasoning and pretrain-stage reasoning.

Research Intern · NExT++ Lab, NUS · Jun 2023 – Jan 2024
Topic: LLM-based agent simulation for recommendation; large-scale LLM agent simulation platform.

Quantitative Research Assistant · CITIC Securities, Research Department · Mar 2021 – Nov 2021
Topic: Fund style classification.