About this episode

This episode discusses how recursive reasoning in AI models can outperform larger models on complex tasks.

A 7-million parameter model outperforming models a thousand times its size on tasks like ARC Prize. That's what recursive reasoning unlocks.In this episode of Decoded, YC's Ankit Gupta and Francois Chaubard break down two recent papers on recursive AI models, HRMs and TRMs, that are achieving state-of-the-art results with a fraction of the parameters of today's largest models.They explain why standard LLMs hit a fundamental ceiling on certain reasoning tasks, how recursion at inference time gives small models the compute depth to break through it, and what happens when you combine these ideas with the power of large-scale foundation models.

People in this episode

Host: Ankit Gupta

Guest: Francois Chaubard

Topics covered

• AI
• recursive reasoning
• scaling laws
• machine learning
• state-of-the-art models

Keywords

• recursive AI
• LLMs
• compute depth
• foundation models
• parameter efficiency

Mentioned in this episode

Organizations: Y Combinator

Books & works: ARC Prize, HRMs, TRMs

Explore listener stats, chart rankings, contacts and more on the Y Combinator Startup Podcast podcast page.