Song Guo

Full Professor, Department of Computer Science and Engineering, Hong Kong University of Science and Technology

3 papers at NeurIPS 2025

Homepage· OpenReview· Semantic Scholar· Google Scholar

Poster Session 5

1 paper

Friday, December 5, 2025 · 11:00 AM → 2:00 PM

Exhibit Hall C,D,E

ICLScan: Detecting Backdoors in Black-Box Large Language Models via Targeted In-context Illumination

#1411 · Xiaoyi Pang, Xuanyi Hao, Song Guo, Qi Luo, Zhibo Wang

This paper reveals that backdoored LLM are more susceptible to new trigger injection during in-context learning (ICL) and proposes a targeted ICL-based lightweight backdoor detection framework for black-box LLMs used for generative tasks.

Poster Session 6

2 papers

Friday, December 5, 2025 · 4:30 PM → 7:30 PM

Exhibit Hall C,D,E

Mesh-RFT: Enhancing Mesh Generation via Fine-grained Reinforcement Fine-Tuning

#4405 Spotlight · Jian Liu, Jing Xu, Song Guo, Jing Li, Guojingfeng, Jiaao Yu, Haohan Weng, Biwen Lei, Xianghui Yang, Zhuo Chen, Fangqi Zhu, Tao Han, Chunchao Guo

We introduce Mesh-RFT, a novel framework leveraging localized preference optimization with topology-aware scoring to achieve state-of-the-art high-fidelity 3D mesh generation.

DiEP: Adaptive Mixture-of-Experts Compression through Differentiable Expert Pruning

#812 · Sikai Bai, Haoxi Li, Jie ZHANG, Zicong Hong, Song Guo