144 citations · 243 across the 64 of their papers we have counts for
13 papers · 1 filter
Archer: Adaptive Reuse of Cached Hidden States for Efficient Rollback in Diffusion Language Models
Xuning He, Zinan Sheng, Qingyuan Shi +5
Diffusion language models (DLMs) can revise earlier predictions as context evolves, but rollback requires costly full-sequence recomputation. Conventional key-value (KV) caching as…
Evaluating the Formal Reasoning Capabilities of Large Language Models through Chomsky Hierarchy
Yihong Dong, Jianha Xiao, Xue Jiang +7
The formal reasoning capabilities of LLMs are crucial for advancing automated software engineering. However, existing benchmarks for LLMs lack systematic evaluation based on comput…
Detecting Data Contamination from Reinforcement Learning Post-training for Large Language Models
Yongding Tao, Tian Wang, Yihong Dong +4
Data contamination poses a significant threat to the reliable evaluation of Large Language Models (LLMs). This issue arises when benchmark samples may inadvertently appear in train…
Rethinking Repetition Problems of LLMs in Code Generation
Yihong Dong, Yuchen Liu, Xue Jiang +2
With the advent of neural language models, the performance of code generation has been significantly boosted. However, the problem of repetitions during the generation process cont…
Theoretical Proof that Auto-regressive Language Models Collapse when Real-world Data is a Finite Set
Lecheng Wang, Xianjie Shi, Ge Li +5
Auto-regressive language models (LMs) have been widely used to generate data in data-scarce domains to train new LMs, compensating for the scarcity of real-world data. Previous wor…
Attr-Int: A Simple and Effective Entity Alignment Framework for Heterogeneous Knowledge Graphs
Linyan Yang, Jingwei Cheng, Chuanhao Xu +3
Entity alignment (EA) refers to the task of linking entities in different knowledge graphs (KGs). Existing EA methods rely heavily on structural isomorphism. However, in real-world…
EvoCodeBench: An Evolving Code Generation Benchmark with Domain-Specific Evaluations
Jia Li, Ge Li, Xuanming Zhang +6
How to evaluate Large Language Models (LLMs) in code generation remains an open question. Existing benchmarks have two limitations - data leakage and lack of domain-specific evalua…
aiXcoder-7B: A Lightweight and Effective Large Language Model for Code Processing
Siyuan Jiang, Jia Li, He Zong +11
Large Language Models (LLMs) have been widely used in code completion, and researchers are focusing on scaling up LLMs to improve their accuracy. However, larger LLMs have lower in…
DevEval: A Manually-Annotated Code Generation Benchmark Aligned with Real-World Code Repositories
Jia Li, Ge Li, Yunfei Zhao +15
How to evaluate the coding abilities of Large Language Models (LLMs) remains an open question. We find that existing benchmarks are poorly aligned with real-world code repositories…
EvoCodeBench: An Evolving Code Generation Benchmark Aligned with Real-World Code Repositories
Jia Li, Ge Li, Xuanming Zhang +2
How to evaluate Large Language Models (LLMs) in code generation is an open question. Existing benchmarks demonstrate poor alignment with real-world code repositories and are insuff…
Generalization or Memorization: Data Contamination and Trustworthy Evaluation for Large Language Models
Yihong Dong, Xue Jiang, Huanyu Liu +4
Recent statements about the impressive capabilities of large language models (LLMs) are usually supported by evaluating on open-access benchmarks. Considering the vast size and wid…
Self-supervised Adaptive Pre-training of Multilingual Speech Models for Language and Dialect Identification
Mohammed Maqsood Shaik, Dietrich Klakow, Badr M. Abdullah
Pre-trained Transformer-based speech models have shown striking performance when fine-tuned on various downstream tasks such as automatic speech recognition and spoken language ide…
PACE: Improving Prompt with Actor-Critic Editing for Large Language Model
Yihong Dong, Kangcheng Luo, Xue Jiang +2
Large language models (LLMs) have showcased remarkable potential across various tasks by conditioning on prompts. However, the quality of different human-written prompts leads to s…