Hi! My name is Jing Xiong (熊璟), and I am currently a first-year PhD student at the University of Hong Kong, supervised by Prof. Yi Huang and Prof. Lingpeng Kong. I have published multiple papers in top-tier conferences and journals including ICLR, ICML, NeurIPS, ACL, EMNLP, and TMLR. My work primarily focuses on efficient language model inference and automated theorem proving.

Previously, I obtained my Master’s degree from the School of Intelligent Systems Engineering, Sun Yat-sen University, under the supervision of Prof. Chengming Li. Before that, I obtained my Bachelor’s degree in Computer Science from Central South University in 2021. I had the honor of collaborating with Prof. Xiaodan Liang.

My Interests

I am interested in research areas related to language modeling, efficient inference, automated theorem proving, and mathematical reasoning.

Service

• Conference reviewer for: NAACL, EMNLP, ACL, ICML, ICLR, NIPS, COLING, etc.

Contact Information

Feel free to reach out to me via email at junexiong@connect.hku.hk or xiongj69@mail2.sysu.edu.cn if you have any questions or would like to collaborate!

News

05/2025: Two paper are accepted in ACL 2025. 05/2025: Two paper are accepted in ICML 2025.
03/2025: A paper is accepted in SLLM Workshop at ICLR 2025. This work aims to bridge this gap by examining the inherent sparsity of standard attention processes.
03/2025: A paper is accepted in TMLR.
01/2025: Two paper are accepted in ICLR 2025.
09/2024: Three paper are accepted in NIPS 2024.
01/2024: A paper is accepted in ICLR 2024.
01/2024: A paper is accepted in ICLR 2024.
11/2023: The paper TRIGO has been selected for an oral presentation at EMNLP 2023.
10/2023: A paper is accepted in EMNLP 2023.
9/2023: A paper has been submitted to ICLR 2024.
9/2023: A paper has been submitted to ICLR 2024.
5/2023: A paper is accepted in ACL 2023.
5/2022: A paper is accepted in SIGIR 2022.

Publications

MEIT: Multi-Modal Electrocardiogram Instruction Tuning on Large Language Models for Report Generation

ACL2025
arXiv, Code

DAPE V2: Process Attention Score as Feature Map for Length Extrapolation
ACL2025

arXiv, Code

ParallelComp: Parallel Long-Context Compressor for Length Extrapolation
ICML2025

arXiv, Code

How Sparse Attention Approximates Exact Attention? Your Attention is Naturally n^C-Sparse
SLLM

arXiv

Autoregressive Models in Vision: A Survey
TMLR

arXiv, Code, Video

FormalAlign: Automated Alignment Evaluation for Autoformalization
ICLR2025

arXiv, Code

D2O: Dynamic Discriminative Operations for Efficient Long-Context Inference of Large Language Models

ICLR2025
arXiv, Code

CAPE: Context-Adaptive Positional Encoding for Length Extrapolation

Nips2024
arXiv, Code

Multimodal large language models make text-to-image generative models align better

Nips2024
arXiv, Code

Boosting text-to-video generative model with MLLMs feedback

Nips2024
arXiv

LEGO-Prover: Neural Theorem Proving with Growing Libraries

ICLR2024
arXiv, Code

DQ-LoRe: Dual Queries with Low Rank Approximation Re-ranking for In-Context Learning

ICLR2024
arXiv, Code

TRIGO: Benchmarking Formal Mathematical Proof Reduction for Generative Language Models

EMNLP2023
arXiv, Code, Video

DT-Solver: Automated Theorem Proving with Dynamic-Tree Sampling Guided by Proof-level Value Function

ACL2023
arXiv

Expression Syntax Information Bottleneck for Math Word Problems

SIGIR2022
arXiv, Code, Video

Self-consistent Reasoning For Solving Math Word Problems

arXiv, Code

CV

Please check my resume here.