Papers with code-related tasks

6 papers
NL-Debugging: Exploiting Natural Language as an Intermediate Representation for Code Debugging (2025.emnlp-main)

Copied to clipboard

Challenge: Early debugging efforts focused on code-level analysis, which often fails when addressing complex programming errors.
Approach: They propose a framework that employs natural language as an intermediate representation to improve code debugging by debuggating at a natural language level.
Outcome: The proposed framework outperforms traditional debugging methods and enables a broader modification space through direct refinement guided by execution feedback.
Surge: On the Potential of Large Language Models as General-Purpose Surrogate Code Executors (2025.emnlp-main)

Copied to clipboard

Challenge: Neural surrogate models are powerful tools in data mining, but are underexplored . large language models (LLMs) have demonstrated remarkable capabilities in code-related tasks .
Approach: They propose a benchmarking framework to examine the feasibility of large language models . they examine scaling laws, data efficiency, and predictive accuracy of 21 open-source and proprietary LLMs .
Outcome: The proposed benchmark examines 21 open-source and proprietary LLMs . it also examines scaling laws, data efficiency, and predictive accuracy .
Read between the lines - Functionality Extraction From READMEs (2024.findings-naacl)

Copied to clipboard

Challenge: While text summarization is a well-known NLP task, functionality extraction from Git README files is based on a novel and useful variant of it.
Approach: They propose a novel task called functionality extraction from Git README files that detects all the functionalities supported by the corresponding application software.
Outcome: The proposed model outperforms baseline models that can be designed using popular black-box or white-box large language models (LLMs) such as ChatGPT and Bard.
DolphCoder: Echo-Locating Code Large Language Models with Diverse and Multi-Objective Instruction Tuning (2024.acl-long)

Copied to clipboard

Challenge: Numerous code large language models (LLMs) have been proposed to enhance code generation performance.
Approach: They propose a diverse instruction model DolphCoder with self-evaluating for code generation that learns diverse instruction targets and combines a code evaluation objective to enhance its code generation ability.
Outcome: The proposed model achieves superior performance on the HumanEval and MBPP benchmarks, demonstrating new insights for future code instruction tuning work.
Alignment with Fill-In-the-Middle for Enhancing Code Generation (2025.emnlp-main)

Copied to clipboard

Challenge: Existing methods for generating test cases with limited training data are not reliable and may be counterproductive.
Approach: They propose a method that splits code snippets into smaller, granular blocks, creating more diverse DPO pairs from the same test cases.
Outcome: The proposed approach shows significant improvements in code generation tasks on benchmark datasets such as HumanEval (+), MBPP (+), and APPS.
HiRoPE: Length Extrapolation for Code Models Using Hierarchical Position (2024.acl-long)

Copied to clipboard

Challenge: Existing LLMs are constrained by their pre-trained context lengths, leading to performance issues . elucidating this limitation, we propose a training-free solution to the context length limitation in LLM applications .
Approach: They propose a method that integrates hierarchical rotary position embedding into LLMs without extra training costs.
Outcome: The proposed method improves performance on language modeling and long code completion tasks.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations