CodeHalu: Investigating Code Hallucinations in LLMs via Execution-based Verification

• URL: http://arxiv.org/abs/2405.00253v3
• Date: Sat, 17 Aug 2024 00:48:05 GMT
• Title: CodeHalu: Investigating Code Hallucinations in LLMs via Execution-based Verification
• Authors: Yuchen Tian, Weixiang Yan, Qian Yang, Xuandong Zhao, Qian Chen, Wen Wang, Ziyang Luo, Lei Ma, Dawn Song,
• Abstract summary: We introduce the concept of code hallucinations and propose a classification method for code hallucination based on execution verification. We present a dynamic detection algorithm called CodeHalu designed to detect and quantify code hallucinations. We also introduce the CodeHaluEval benchmark, which includes 8,883 samples from 699 tasks, to systematically and quantitatively evaluate code hallucinations.
• Score: 73.66920648926161
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Language Models (LLMs) have made significant progress in code generation, offering developers groundbreaking automated programming support. However, LLMs often generate code that is syntactically correct and even semantically plausible, but may not execute as expected or fulfill specified requirements. This phenomenon of hallucinations in the code domain has not been systematically explored. To advance the community's understanding and research on this issue, we introduce the concept of code hallucinations and propose a classification method for code hallucination based on execution verification. We categorize code hallucinations into four main types: mapping, naming, resource, and logic hallucinations, with each category further divided into different subcategories to understand and address the unique challenges faced by LLMs in code generation with finer granularity. Additionally, we present a dynamic detection algorithm called CodeHalu designed to detect and quantify code hallucinations. We also introduce the CodeHaluEval benchmark, which includes 8,883 samples from 699 tasks, to systematically and quantitatively evaluate code hallucinations. By evaluating 17 popular LLMs using this benchmark, we reveal significant differences in their accuracy and reliability in code generation, offering detailed insights for further improving the code generation capabilities of LLMs. The CodeHalu benchmark and code are publicly available at https://github.com/yuchen814/CodeHalu.

Related papers

• MLLM can see? Dynamic Correction Decoding for Hallucination Mitigation [50.73561815838431]
  Multimodal Large Language Models (MLLMs) frequently exhibit hallucination phenomena. We propose a novel dynamic correction decoding method for MLLMs (DeCo) We evaluate DeCo on widely-used benchmarks, demonstrating that it can reduce hallucination rates by a large margin compared to baselines.
  arXiv  Detail & Related papers  (2024-10-15T16:57:44Z)
• Collu-Bench: A Benchmark for Predicting Language Model Hallucinations in Code [20.736888384234273]
  We introduce Collu-Bench, a benchmark for predicting code hallucinations of large language models (LLMs) Collu-Bench includes 13,234 code hallucination instances collected from five datasets and 11 diverse LLMs, ranging from open-source models to commercial ones. We conduct experiments to predict hallucination on Collu-Bench, using both traditional machine learning techniques and neural networks, which achieves 22.03 -- 33.15% accuracy.
  arXiv  Detail & Related papers  (2024-10-13T20:41:47Z)
• LongHalQA: Long-Context Hallucination Evaluation for MultiModal Large Language Models [96.64960606650115]
  LongHalQA is an LLM-free hallucination benchmark that comprises 6K long and complex hallucination text. LongHalQA is featured by GPT4V-generated hallucinatory data that are well aligned with real-world scenarios.
  arXiv  Detail & Related papers  (2024-10-13T18:59:58Z)
• LLM Hallucinations in Practical Code Generation: Phenomena, Mechanism, and Mitigation [33.46342144822026]
  Code generation aims to automatically generate code from input requirements, significantly enhancing development efficiency. Recent large language models (LLMs) based approaches have shown promising results and revolutionized code generation task. Despite the promising performance, LLMs often generate contents with hallucinations, especially for the code generation scenario.
  arXiv  Detail & Related papers  (2024-09-30T17:51:15Z)
• CodeMirage: Hallucinations in Code Generated by Large Language Models [6.063525456640463]
  Large Language Models (LLMs) have shown promising potentials in program generation and no-code automation. LLMs are prone to generate hallucinations, i.e., they generate text which sounds plausible but is incorrect. We propose the first benchmark CodeMirage dataset for code hallucinations.
  arXiv  Detail & Related papers  (2024-08-14T22:53:07Z)
• Uncovering LLM-Generated Code: A Zero-Shot Synthetic Code Detector via Code Rewriting [78.48355455324688]
  We propose a novel zero-shot synthetic code detector based on the similarity between the code and its rewritten variants. Our results demonstrate a notable enhancement over existing synthetic content detectors designed for general texts.
  arXiv  Detail & Related papers  (2024-05-25T08:57:28Z)
• Exploring and Evaluating Hallucinations in LLM-Powered Code Generation [14.438161741833687]
  Large Language Models (LLMs) produce outputs that deviate from users' intent, exhibit internal inconsistencies, or misalign with factual knowledge. Existing work mainly focuses on investing the hallucination in the domain of natural language generation. We conduct a thematic analysis of the LLM-generated code to summarize and categorize the hallucinations present in it. We propose HalluCode, a benchmark for evaluating the performance of code LLMs in recognizing hallucinations.
  arXiv  Detail & Related papers  (2024-04-01T07:31:45Z)
• InfiBench: Evaluating the Question-Answering Capabilities of Code Large Language Models [56.723509505549536]
  InfiBench is the first large-scale freeform question-answering (QA) benchmark for code to our knowledge. It comprises 234 carefully selected high-quality Stack Overflow questions that span across 15 programming languages. We conduct a systematic evaluation for over 100 latest code LLMs on InfiBench, leading to a series of novel and insightful findings.
  arXiv  Detail & Related papers  (2024-03-11T02:06:30Z)
• Code Prompting Elicits Conditional Reasoning Abilities in Text+Code LLMs [65.2379940117181]
  We introduce code prompting, a chain of prompts that transforms a natural language problem into code. We find that code prompting exhibits a high-performance boost for multiple LLMs. Our analysis of GPT 3.5 reveals that the code formatting of the input problem is essential for performance improvement.
  arXiv  Detail & Related papers  (2024-01-18T15:32:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.