Stabilizing Recurrent Dynamics for Test-Time Scalable Latent Reasoning in Looped Language Models.
Authors & venue
The 43rd International Conference on Machine Learning (ICML 2026).
Xiao-Wen Yang杨骁文
Xiaohongshu · AllSpark
Research Intern · Agent post-training · Jun 2026–present
Ph.D. Student · Nanjing University
LAMDA · Advisor: Yu-Feng Li
纪元总览
NANJING, CHINA↓ PUBLICATIONS
News
Our knowledge augmentation paper is published at KDD 2026.
FormalImG accepted to Findings of EMNLP 2026.
Earlier news +
Our IJCAI Travel Planning Challenge review is published in Frontiers of Computer Science.
STARS and Pianist Transformer accepted to ICML 2026.
FormalML and ChinaTravel accepted to ICLR 2026.
Self-Backtracking accepted to AAAI 2026.
Our neuro-symbolic reasoning survey accepted to IJCAI 2025.
CARTS accepted to ICLR 2025.
Publications.
纪元总览
文明档案* Equal contribution† Corresponding author
Mitigating Neuro-Symbolic Reasoning Shortcuts with Data-Driven Knowledge Augmentation.
Authors & venue
The 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2026), V.2, pp. 2861–2872.
Pianist Transformer: Towards Expressive Piano Performance Rendering via Scalable Self-Supervised Pre-Training.
Authors & venue
The 43rd International Conference on Machine Learning (ICML 2026).
FormalImG: Evaluating Structural Compositional Generalization for T2I Models.
Authors & venue
Findings of EMNLP 2026 (accepted). Earlier version: ICML 2026 Workshop on Combining Theory and Benchmarks.
A Review of the IJCAI 2025 Travel Planning Challenge: Bench-marking LLM Agents with Compositional Constraints.
Authors & venue
Frontiers of Computer Science, 2026. Published online June 12, 2026.
Re²: Reflective Rule Induction and Rule-Guided Refinement for Embodied Planning.
Authors & venue
The 2nd Workshop on Foundation Models Meet Embodied Agents at CVPR 2026.
FormalML: A Benchmark for Evaluating Formal Subgoal Completion in Machine Learning Theory.
Authors & venue
The 14th International Conference on Learning Representations (ICLR 2026).
ChinaTravel: An Open-Ended Benchmark for Language Agents in Chinese Travel Planning.
Authors & venue
The 14th International Conference on Learning Representations (ICLR 2026).
Step Back to Leap Forward: Self-Backtracking for Symbolic Reasoning and Planning in Language Models.
Authors & venue
The 40th AAAI Conference on Artificial Intelligence (AAAI 2026).
Programmatic Context Augmentation for LLM-based Symbolic Regression.
Authors & venue
Preprint, 2026.
Thinking with Tables: Enhancing Multi-Modal Tabular Understanding via Neuro-Symbolic Reasoning.
Authors & venue
Preprint, 2026.
Hindsight Credit Assignment for Long-Horizon LLM Agents.
Authors & venue
Preprint, 2026.
Neuro-Symbolic Artificial Intelligence: Towards Improving the Reasoning Abilities of Large Language Models.
Authors & venue
The 34th International Joint Conference on Artificial Intelligence (IJCAI 2025 Survey Track).
CARTS: Advancing Neural Theorem Proving with Diversified Tactic Calibration and Bias-Resistant Tree Search.
Authors & venue
The 13th International Conference on Learning Representations (ICLR 2025).
Analysis for Abductive Learning and Neural-Symbolic Reasoning Shortcuts.
Authors & venue
The 41st International Conference on Machine Learning (ICML 2024).
Safe Abductive Learning in the Presence of Inaccurate Rules.
Authors & venue
The 38th AAAI Conference on Artificial Intelligence (AAAI 2024).
AutoEvolve: Automatically Evolving Queries for Applicable and Scalable Retrieval-Augmented Generation Benchmarking.
Authors & venue
The 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025).
Curriculum Abductive Learning for Mitigating Reasoning Shortcuts.
Authors & venue
The 34th International Joint Conference on Artificial Intelligence (IJCAI 2025).
D3: Diversity, Difficulty, and Dependability-Aware Data Selection for Sample-Efficient LLM Instruction Tuning
Authors & venue
The 34th International Joint Conference on Artificial Intelligence (IJCAI 2025).
Abductive Learning for Neuro-Symbolic Grounded Imitation.
Authors & venue
The 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD’25).
LOG: Active Model Adaptation for Label-Efficient OOD Generalization.
Authors & venue
The 36th Conference on Neural Information Processing Systems (NeurIPS 2022).
Open-Set Learning under Covariate Shift.
Authors & venue
Machine Learning (ML’22).
LawGPT: Knowledge-Guided Data Generation and Its Application to Legal LLM.
Authors & venue
Preprint, 2025.
Investigating the Limitation of CLIP Models: The Worst-Performing Categories.
Authors & venue
Preprint, 2023.
Lightweight Retrieval Tuning for Black-Box Language Models.
Authors & venue
NeurIPS’23 ENSLP Workshop.
No papers found.
More about me +
Ph.D. student at Nanjing University, advised by Yu-Feng Li, in the LAMDA Group led by Zhi-Hua Zhou.
Research Intern at Xiaohongshu AllSpark since June 2026, working on agent post-training.
B.Sc. in Computer Science, Nanjing University, 2023. Research: machine learning and artificial general intelligence.
Software
- LawGPT: A Large Language Model in the Legal Domain. [GitHub]
Honor
- National Scholarship, Nanjing University, Oct 2024
- Outstanding Student, Nanjing University, Nov 2024
- President’s Scholarship, Nanjing University, Oct 2023
- Outstanding Graduate, Nanjing University, Jun 2023
Academic Service
Reviewer for:
- AAAI 2024, 2025, 2026
- ICLR 2024, 2025, 2026
- ICML 2024, 2025, 2026
- NeurIPS 2023, 2024, 2025
Meta Reviewer for IJCAI 2025