Rishabh Agarwal

Chronological feed of everything captured from Rishabh Agarwal.

paper / agarwl_ / May 16 / failed

Learning, Fast and Slow: Towards LLMs That Adapt Continually

paper / agarwl_ / May 13 / failed

Towards Compute-Optimal Many-Shot In-Context Learning

paper / agarwl_ / May 13 / failed

The Art of Scaling Reinforcement Learning Compute for LLMs

paper / agarwl_ / May 13 / failed

Shape of Thought: When Distribution Matters More than Correctness in Reasoning Tasks