Chain-of-illocution prompting improves source adherence in RAG explanations for programming education by up to 63% over baselines.
arXiv preprint arXiv:2402.08801 (2024)
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
abstract
Since its release in November 2022, ChatGPT has shaken up Stack Overflow, the premier platform for developers queries on programming and software development. Demonstrating an ability to generate instant, human-like responses to technical questions, ChatGPT has ignited debates within the developer community about the evolving role of human-driven platforms in the age of generative AI. Two months after ChatGPT release, Meta released its answer with its own Large Language Model (LLM) called LLaMA: the race was on. We conducted an empirical study analyzing questions from Stack Overflow and using these LLMs to address them. This way, we aim to (i) quantify the reliability of LLMs answers and their potential to replace Stack Overflow in the long term; (ii) identify and understand why LLMs fail; (iii) measure users activity evolution with Stack Overflow over time; and (iv) compare LLMs together. Our empirical results are unequivocal: ChatGPT and LLaMA challenge human expertise, yet do not outperform it for some domains, while a significant decline in user posting activity has been observed. Furthermore, we also discuss the impact of our findings regarding the usage and development of new LLMs and provide guidelines for future challenges faced by users and researchers.
representative citing papers
A single vulnerable Node.js path traversal pattern was found in 1,756 GitHub projects, most rated critical, and the authors' automated pipeline produced patches, disclosures, and evidence that LLMs have learned the pattern.
DOVE measures LLM cultural value alignment via a rate-distortion value codebook and unbalanced optimal transport between human and model open-ended text distributions.
A 31-developer survey found that LLM-generated explanations of GitHub Actions failures are perceived as correct and clear for simple logs, but less useful for complex CI/CD failures.
GENCNIPPET is a planned browser plugin that would use a fine-tuned Llama-3 model to generate code snippets for Stack Overflow questions missing them, with evaluation plans but no implemented results.
Empirical study of Stack Overflow logging posts identifies 11 topics where containerized environments show the highest difficulty via unanswered rates and resolution time.
citing papers explorer
-
Illocutionary Explanation Planning for Source-Faithful Explanations in Retrieval-Augmented Language Models
Chain-of-illocution prompting improves source adherence in RAG explanations for programming education by up to 63% over baselines.
-
Eradicating the Unseen: Detecting, Exploiting, and Remediating a Path Traversal Vulnerability across GitHub
A single vulnerable Node.js path traversal pattern was found in 1,756 GitHub projects, most rated critical, and the authors' automated pipeline produced patches, disclosures, and evidence that LLMs have learned the pattern.
-
Distributional Open-Ended Evaluation of LLM Cultural Value Alignment Based on Value Codebook
DOVE measures LLM cultural value alignment via a rate-distortion value codebook and unbalanced optimal transport between human and model open-ended text distributions.
-
Explaining GitHub Actions Failures with Large Language Models: Challenges, Insights, and Limitations
A 31-developer survey found that LLM-generated explanations of GitHub Actions failures are perceived as correct and clear for simple logs, but less useful for complex CI/CD failures.
-
GENCNIPPET: Automated Generation of Code Snippets for Supporting Programming Questions
GENCNIPPET is a planned browser plugin that would use a fine-tuned Llama-3 model to generate code snippets for Stack Overflow questions missing them, with evaluation plans but no implemented results.
-
An Empirical Study on Logging Evolution On Stack Overflow: Trends, Topics, and Challenges
Empirical study of Stack Overflow logging posts identifies 11 topics where containerized environments show the highest difficulty via unanswered rates and resolution time.