Hi! I am a PhD student at LMU Munich supervised by Prof. Dr. Volker Tresp. Prior to this, I obtained my M.Sc. in Data Engineering and Analytics at the Technical University of Munich, and my B.E. in Computer Science and Technology at Xi'an Jiaotong University.
My research interests lie in understanding the capabilities and limitations of language models, and from there developing new methods and solutions. I am currently working on projects related to recursive self-improvement, architectural designs, and mechanistic interpretability of language models. We are open to new collaboration opportunities. Please feel free to reach out if you are interested.

Selected Publications

Mendel Gödel Machine Polyglot performance figure
Changzhi Liu*, Yilun Liu*, Sikuan Yan, Volker Tresp, Yunpu Ma (*equal contribution)
arXiv preprint [Demo][Paper][Code]
  • Self-improving agents with additional comparative signals derived from the archive of past attempts
  • Better performance and efficiency, with 35B open models outperforming 117× larger frontier models
  • Generalizablity across benchmarks and models, yielding genuinely reusable and scalable scaffolds
binomial ordering preference figure
Zhiqing Yang*, Yilun Liu*, Yunpu Ma, Volker Tresp, Hinrich Schütze (*equal contribution)
arXiv preprint [Paper][Code]
  • Evaluation protocol of binomial ordering preference for LLMs, with a dataset of 600 binomial pairs across 8 languages
  • LLMs behaviorally align with the empirically preferred direction more reliably than the strength, and that strength can be representationally located and manipulated
Fiberwise optimal transport schedule overview figure
Luyi Jia, Boyan Zhang, Yilun Liu, Steffen Rulands
arXiv preprint [Paper]
  • Model-aware schedule construction via fiberwise optimal transport, combining prediction risk with kinetic action for closed-form time allocation, yielding consistent gains over strong baselines
  • Normalized risk profiles and schedule deformations align across models, with a frozen analytic form retaining most gains
Routing-Free MoE overview figure
Yilun Liu*, Jinru Han*, Sikuan Yan, Volker Tresp, Yunpu Ma (*equal contribution)
EMNLP 2026 Main Conference [Paper][Code]
  • Routing-Free MoE architecture, eliminating routers, Softmax, TopK, and hard-coded load balancing
  • Unified, adaptive load-balancing framework jointly optimizing token- and expert-balancing
  • Improved performance and efficiency over baselines
PyraVid hierarchical multimodal memory figure
Sikuan Yan, Sicheng Dong, Haotong Wang, Ercong Nie, Yilun Liu, Jinhe Bi, Yingjie Xu, Susanna Schwarzmann, Riccardo Trivisonno, Volker Tresp, Yunpu Ma
EMNLP 2026 Main Conference [Paper]
  • Hierarchical multimodal memory with a coarse-to-fine pyramid structure for long-horizon video reasoning
  • Structure-guided memory expansion with pruning for causally related but semantically distant events
  • Consistent gains across long-video benchmarks, model scales, and question types
SIREN overview figure
Difan Jiao, Yilun Liu, Ye Yuan, Zhenwei Tang, Linfeng Du, Haolun Wu, Ashton Anderson
ACL 2026 Main Conference [Paper][Code][Poster]
  • SIREN, a plug-and-play component harnessing LLM internal representations for harmfulness detection
  • Outperforming dedicated safety guardrails in performance, generalization, and efficiency
PERFT performance figure
Yilun Liu*, Yunpu Ma*, Shuo Chen, Zifeng Ding, Bailan He, Zhen Han, Volker Tresp (*equal contribution)
EACL 2026 Findings [Paper][Code][Poster]
  • Dynamics between memory vectors in experts and expert vectors in routers when PEFTing to MoE LLMs
  • Unified framework for integrating PEFT with MoE LLMs
  • PERFT, a family of effective and scalable adaptation strategies
SPIN aggregated feature analysis figure
Difan Jiao*, Yilun Liu*, Zhenwei Tang, Danial Matter, Jürgen Pfeffer, Ashton Anderson (*equal contribution)
ACL 2024 Findings [Paper][Code][Poster]
  • Lightweight plug-and-play text classification frameworkusing LLM internal representations
  • Superior performance, efficiency, and interpretability compared to conventional methods

Educational Background

10.2025 – present
Ph.D. Student (in progress)
10.2021 –
M.Sc. Data Engineering and Analytics
Department of Informatics, Technical University of Munich
Munich, Bavaria, Germany
09.2017 – 07.2021
B.E. Computer Science and Technology
Faculty of Electronic and Information Engineering, Xi'an Jiaotong University
Xi'an, Shaanxi, China
09.2016 – 07.2017
Pre-university Education, Honors Youth Program
Qian Xuesen Honors College, XJTU & Tianjin Nankai High School
Xi'an, Shaanxi; Tianjin, China

Academic Service

Reviewer ICLR (2025, 2026) · COLM (2025, 2026) · ACL Rolling Review (multiple cycles) · IC2S2 (2025)