Pulkit Agrawal

RL's Razor: Why Online Reinforcement Learning Forgets Less

Large Pretraining Datasets Don't Guarantee Robustness after Fine-Tuning

Self-Adapting Language Models

Emergence and Effectiveness of Task Vectors in In-Context Learning: An Encoder Decoder Perspective

Diffusion Policy Policy Optimization

Language Model Personalization via Reward Factorization

From Context to Concept: Concept Encoding in In-Context Learning

ORSO: Accelerating Reward Design via Online Reward Selection and Policy Optimization

Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning

ImageNet-RIB Benchmark: Large Pre-Training Datasets Don't Guarantee Robustness after Fine-Tuning

Collective Model Intelligence Requires Compatible Specialization