Irina Rish

This article discusses improving VLMs' resistance to adversarial attacks through design choices.

Analyzing how VPT, an AI agent, makes decisions in Minecraft gameplay.

This paper highlights the performance of ternary language models and their efficiency.

Examining how computer vision models can align with human visual understanding.

New techniques enhance large language models' ability in complex arithmetic reasoning.

RedPajama datasets aim to enhance language model training through transparency and quality data.

Learn how asynchronous techniques improve real-time decision-making for AI agents.