FlexiViT: One Model for All Patch Sizes
Paper • 2212.08013 • Published • 1
Pavel Izmailov
pavelizmailov
·
AI & ML interests
LLMs, Reasoning, Reinforcement Learning, AI for Science, AI for Math, Alignment, World Models
Recent Activity
authored a paper 4 days ago
FlexiViT: One Model for All Patch Sizes authored a paper 4 days ago
Does Knowledge Distillation Really Work? authored a paper 4 days ago
OpenAI o1 System CardOrganizations
authored 9
papers 4 days ago
Does Knowledge Distillation Really Work?
Paper • 2106.05945 • Published
OpenAI o1 System Card
Paper • 2412.16720 • Published • 37
Closing the Train-Test Gap in World Models for Gradient-Based Planning
Paper • 2512.09929 • Published
From Entropy to Epiplexity: Rethinking Information for Computationally Bounded Intelligence
Paper • 2601.03220 • Published • 2
Reliable and Responsible Foundation Models: A Comprehensive Survey
Paper • 2602.08145 • Published • 8
Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations
Paper • 2204.02937 • Published
Averaging Weights Leads to Wider Optima and Better Generalization
Paper • 1803.05407 • Published • 2
When Can LLMs Learn to Reason with Weak Supervision?
Paper • 2604.18574 • Published • 24