This repository addresses the challenge of staying current with recent research in adversarial learning and adversarial examples, particularly following the unavailability of a previous comprehensive resource. It serves as a curated collection of the latest papers for researchers, engineers, and practitioners in the field, providing a centralized and up-to-date reference point.

How It Works

The project functions by aggregating and maintaining a list of recent academic papers focused on adversarial learning. The core approach involves systematically collecting and presenting these papers, with an emphasis on recency and coverage of transfer-based attacks. This curated list aims to provide a valuable, up-to-date resource for the research community by filling a gap left by a previously unavailable resource.

Quick Start & Requirements

• This repository contains a list of research papers and does not involve executable code, thus requiring no installation or specific runtime environment.
• No prerequisites or setup are necessary to access the paper list.
• Links: The repository itself is the primary resource.

Highlighted Details

• Features a comprehensive list of recent papers on adversarial examples and transfer-based attacks.
• Includes papers dated from November 2025 down to May 2025, aiming for cutting-edge research. (Note: The presence of future dates may indicate an error in the source text.)
• Incorporates data from the "List of All Adversarial Example Papers" up to September 1, 2023.
• Actively seeks community contributions for missed papers via email.

Maintenance & Community

• Maintenance appears to be driven by the repository's creator(s), with a call for community submissions of overlooked papers.
• No specific community channels (Discord, Slack) or roadmap links are provided.

Licensing & Compatibility

• No license information is specified in the provided README content.
• Compatibility notes are not applicable as the repository is a list of papers.

Limitations & Caveats

• The paper list is explicitly stated to be not exhaustive ("may not encompass every paper").
• The reliance on a curated list means it's dependent on the maintainers' efforts and may not capture all emerging research.
• The presence of future dates (2025) for papers suggests a potential issue with the data source or its interpretation, impacting the perceived recency and accuracy of the listed research.