SOTAVerified

Code Documentation Generation

Code Documentation Generation is a supervised task where a code function is the input to the model, and the model generates the documentation for this function.

Description from: CodeTrans: Towards Cracking the Language of Silicone's Code Through Self-Supervised Deep Learning and High Performance Computing

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–8 of 8 papers

TitleStatusHype
DocAgent: A Multi-Agent System for Automated Code Documentation GenerationCode3
RepoAgent: An LLM-Powered Open-Source Framework for Repository-level Code Documentation GenerationCode4
A Comparative Analysis of Large Language Models for Code Documentation Generation—0
Assemble Foundation Models for Automatic Code SummarizationCode1
Memorization and Generalization in Neural Code Intelligence ModelsCode0
CodeTrans: Towards Cracking the Language of Silicon's Code Through Self-Supervised Deep Learning and High Performance ComputingCode1
HAConvGNN: Hierarchical Attention Based Convolutional Graph Neural Network for Code Documentation Generation in Jupyter NotebooksCode0
CodeBERT: A Pre-Trained Model for Programming and Natural LanguagesCode1
Show:102550
All datasetsCodeSearchNet - JavaCodeSearchNet - JavaScriptCodeSearchNet - PhpCodeSearchNetCodeSearchNet - GoCodeSearchNet - PythonCodeSearchNet - Ruby

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1TransformerSmoothed BLEU-425.61—Unverified
2CodeTrans-TF-LargeSmoothed BLEU-418.98—Unverified
3CodeBERT (MLM+RTD)Smoothed BLEU-49.54—Unverified
4CodeBERT (RTD)Smoothed BLEU-48.73—Unverified
5CodeBERT (MLM)Smoothed BLEU-48.51—Unverified
6pre-train w/ code onlySmoothed BLEU-48.3—Unverified
7seq2seqSmoothed BLEU-46.88—Unverified
8RoBERTaSmoothed BLEU-45.72—Unverified