Source Attribution
Source attribution in artificial intelligence focuses on identifying the origin of generated content, whether text, images, music, or other data, and verifying its factual accuracy. Current research emphasizes developing robust methods for detecting and attributing AI-generated content using techniques like contrastive learning, vision-language models, and watermarking, often within specific domains such as climate science or media forensics. This field is crucial for addressing concerns about misinformation, copyright infringement, and ensuring the trustworthiness and accountability of AI systems, impacting both scientific integrity and practical applications.
Papers
The Data Provenance Initiative: A Large Scale Audit of Dataset Licensing & Attribution in AI
Shayne Longpre, Robert Mahari, Anthony Chen, Naana Obeng-Marnu, Damien Sileo, William Brannon, Niklas Muennighoff, Nathan Khazam, Jad Kabbara, Kartik Perisetla, Xinyi Wu, Enrico Shippole, Kurt Bollacker, Tongshuang Wu, Luis Villa, Sandy Pentland, Sara Hooker
Sum-of-Parts: Faithful Attributions for Groups of Features
Weiqiu You, Helen Qu, Marco Gatti, Bhuvnesh Jain, Eric Wong