Zhiren Gong | Academic Homepage

News

May 2026

Two papers accepted by ICML: SubspacePath Pruner and XDomainBench.

Jan 2026

One paper accepted by ICLR: Rule-Guided Active Inference.

Aug 2025

Joined NTU as a PhD candidate and received the NTU Research Scholarship (RSS).

Feb 2025

Joined CUHKSZ as a research intern.

Selected Publications

SubspacePath Pruner: Inference-time Pruning via Probe-based Representation-Parameter Coupling

Zhiren Gong, Yikun Hou, Fan Wu, Che Wang, Fuyao Zhang, Tiantong Wu, Yurong Hao, Jiaming Zhang, Yiyang Duan, Tiantong Wang, Fei Huang, Chau Yuen, Wei Yang Bryan Lim

ICML 2026

Paper Project Page Code

SubspacePath Pruner introduces an inference-time pruning mechanism that couples representation probes with parameter pathways, enabling scenario-adaptive routing that retains robustness while reducing unnecessary computation.

# Model Efficiency # Inference-Time Pruning # Routing

Learning Human Habits with Rule-Guided Active Inference

Zhiren Gong, Chao Yang, Wendi Ren, Shuang Li

ICLR 2026

Paper Project Page Code

This work combines active inference with explicit rule constraints to model habit dynamics in sequential decision settings, improving action quality and latency in realistic environments.

# Active Inference # Decision Intelligence # Interpretable Planning

State of Thought Enables Endogenous Reasoning

Zhiren Gong, Yikun Hou, Zeng Zihao, Ming Xiao, Chau Yuen, Wei Yang Bryan Lim

Coming soon

Paper Project Page Code

SoT studies endogenous state-conditioned reasoning to improve quality-efficiency trade-offs across backbones and benchmark families.

# Reasoning # Efficiency # State Modeling

LoTR: Logic-of-Thought Routing for Plug-and-Play Reasoning of LLMs

Zhiren Gong, Ming Xiao, Chau Yuen, Wei Yang Bryan Lim

Coming soon

Paper Project Page Code

LoTR enhances existing external reasoning paradigms by routing internal computation pathways according to evolving logic states at each reasoning step.

# Plug-and-Play # Logic Routing # LLM Reasoning

Why Mixture of Experts Needs Fewer Samples Than Dense Networks: The Information Exponent Collapse Theorem

Tiantong Wang, Zhiren Gong, Yikun Hou, Tiantong Wu, Wei Yang Bryan Lim

Coming soon

Paper Project Page Code

This paper provides a theory for why sparse MoE architectures can reduce sample complexity barriers compared with dense models via geometry-aware routing and exponent-collapse behavior.

# Mixture of Experts # Sample Complexity # Theory

XDomainBench: Diagnosing Reasoning Collapse in High-Dimensional Scientific Knowledge Composition

Zhiren Gong, Tiantong Wu, Jiaming Zhang, Fuyao Zhang, Che Wang, Yurong Hao, Yikun Hou, Foo Ping, Yilei Zhao, Fei Huang, Chau Yuen, Wei Yang Bryan Lim

ICML 2026

Paper Project Page Code

XDomainBench provides a systematic benchmark for stress-testing large models under cross-domain scientific composition, revealing where and why reasoning collapses as composition complexity increases.

# Benchmark # Reliability # Scientific Reasoning

C2U in RAG: Compositional Concept Unlearning in Retrieval-Augmented Generation

Yiyang Duan, Fan Wu, Tiantong Wu, Che Wang, Zhiren Gong, Wei Guo, Yang Cao, Wei Yang Bryan Lim

Coming soon

Paper Project Page Code

C2U in RAG studies compositional concept unlearning in retrieval-augmented generation to selectively remove target knowledge while preserving utility on retained concepts.

# RAG # Unlearning # Compositionality

Channel Gains to Captions: Task-Unified Multi-Level RF Sensing with Vision-Language Models

Tianyu Hu, Zhiren Gong, Haowei Cui, Shuai Wang, Samson Lasaulce, Lingxiang Li, Wassim Hamidouche, Zhi Chen, Merouane Debbah

Coming soon

Paper Project Page Code

This work studies task-unified RF sensing with vision-language models, aiming to align channel-level signal structures with caption-level semantic outputs in a single framework.

# RF Sensing # Vision-Language Models # Multimodal Intelligence

3D Radio Map Reconstruction based on Generative Adversarial Networks under Constrained Aircraft Trajectories

Tianyu Hu, Yang Huang, Junting Chen, Qihui Wu, Zhiren Gong

IEEE TVT 2023

Paper PDF Code

A GAN-based method reconstructs 3D radio maps from sparse aircraft measurements under trajectory constraints, improving reconstruction quality under realistic sensing conditions.

# Wireless AI # 3D Radio Map # GAN

When Large Language Models Meet the Physical World: Holistic World Intelligence Collapses

Yang Liu, Zhiren Gong, Xiaoping Wang, Jianbo Zheng, Guanghui Ye, Jinhong Hu, Kai Lu, Wei Yang Bryan Lim

Coming soon

Paper Project Page Code

This work analyzes capability collapse when LLM-style systems face physical-world, non-semantic modalities, and studies how to build safer and more robust cognition in such settings.

# LLM + Physical World # Robustness # Reliability

Honors

NTU Research Scholarship (RSS), PhD Program.
NUAA Outstanding Graduate (Top 1%) and Outstanding Thesis 2023 (Top 4%).
Headmaster's Special Mention (2020, 2022), highest undergraduate honor at NUAA.
Top 10 Outstanding Youth of NUAA (2022).
Aviation Industry Honors Scholarship (Top 0.01%).
National Scholarship of China (Top 1%).
First Prize Outstanding Student Scholarship (Top 5%, 2019 - 2023).

Service

Reviewer: ICML 2026 (Gold Reviewer), NeurIPS 2026, ICIC 2026, ICML AdaptFM Workshop.

Teaching Assistant: Data Science Foundations (SC3021), NTU, 2026 Spring.

Academic Talk: ICML 2026.