Navigating Papers With Code
Learn to efficiently search, filter, and discover papers with implementations. Master the site's navigation to find exactly what you need in seconds.
The Homepage
The Papers With Code homepage gives you immediate access to the most important sections:
- Trending Papers: Papers gaining the most attention and GitHub stars right now
- Greatest Papers: All-time most popular papers on the platform
- Latest Papers: Newly indexed papers from arXiv and conferences
- Top Navigation: Quick links to Papers, Datasets, Methods, and SOTA sections
Search Strategies
The search bar is the fastest way to find specific papers or topics. Here are effective search strategies:
Keyword Search
Search by topic, technique name, or problem type. Examples:
vision transformer— finds papers about ViT and related architecturestext-to-image diffusion— finds papers on diffusion-based image generationretrieval augmented generation— finds RAG papers and implementations
Author Search
Search for papers by specific researchers to find their body of work and implementations.
Conference Filtering
Filter papers by venue (NeurIPS, ICML, ICLR, CVPR, ACL, EMNLP) to find peer-reviewed, high-quality research.
Paper Pages
Each paper page on the platform contains rich information:
- Abstract: The paper's summary and key contributions
- Code repositories: Links to all known implementations, ranked by GitHub stars
- Results: Benchmark scores reported in the paper, linked to SOTA leaderboards
- Tasks: The ML tasks the paper addresses (e.g., object detection, text classification)
- Datasets: Datasets used for training or evaluation
- Methods: Specific techniques and components used in the paper
Trending and Discovery
Papers With Code offers several ways to discover new and impactful research:
Trending Papers
Updated daily, trending papers are ranked by a combination of recent GitHub activity, social media mentions, and community engagement. This is the best way to stay on top of breaking research.
Newsletter
Subscribe to the weekly newsletter to receive curated highlights of the most important papers and developments directly in your inbox.
Task Pages
Browse by task (e.g., Image Classification, Machine Translation, Question Answering) to see all papers, benchmarks, and datasets related to a specific problem.
Using the API
Papers With Code also offers a REST API for programmatic access to papers, datasets, benchmarks, and results. This is useful for building research tools, automated literature reviews, or integration with your own workflows.