FeatX extracts epic-feature hierarchies with code mappings from repositories and applies feature edits via a three-stage Evolution Agent, reporting 42.6% relative F1 gain in function-level localization and lower cognitive load versus vanilla ChatGPT in a user study and 38-commit replay.
Reposummary: Feature-oriented summarization and documentation generation for code repositories
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.SE 4years
2026 4verdicts
UNVERDICTED 4roles
background 1polarities
background 1representative citing papers
A clustering-based pipeline generates individual and integration-level test specifications from thousands of automotive requirements by grouping embeddings, summarizing clusters, and applying LLM calls with bounded context and standards grounding.
Single-agent RAG pipeline matches multi-agent lexical quality for README generation while cutting token consumption by 86% and doubling speed, with developer-guided planning yielding the highest overall quality.
A systematic review that categorizes prompting strategies for LLM-based code summarization, assesses their effectiveness, and identifies gaps in research and evaluation practices.
citing papers explorer
-
FeatX: Editing Software by Editing Features for Repository-Level Code Evolution
FeatX extracts epic-feature hierarchies with code mappings from repositories and applies feature edits via a three-stage Evolution Agent, reporting 42.6% relative F1 gain in function-level localization and lower cognitive load versus vanilla ChatGPT in a user study and 38-commit replay.
-
Cluster-Aware Dual-Level Test Specification Generation for Large-Scale Automotive Software Requirements
A clustering-based pipeline generates individual and integration-level test specifications from thousands of automotive requirements by grouping embeddings, summarizing clusters, and applying LLM calls with bounded context and standards grounding.
-
The Illusion of Agentic Complexity in README.md Generation: Evaluating Single-Agent vs. Multi-Agent RAG Systems
Single-agent RAG pipeline matches multi-agent lexical quality for README generation while cutting token consumption by 86% and doubling speed, with developer-guided planning yielding the highest overall quality.
-
Prompt-Driven Code Summarization: A Systematic Literature Review
A systematic review that categorizes prompting strategies for LLM-based code summarization, assesses their effectiveness, and identifies gaps in research and evaluation practices.