factual

RL Episodes Yield Only O(1) Bits

An LLM being RLed on outcome-based rewards learns only O(1) bits per episode even though the episode may be tens of thousands of tokens long, whereas animals extract far more information from interacting with their environment than just the end-of-episode reward signal.

factualpending

Speaker

Dwarkesh Patel

Evidence Quote

“We animals clearly extract far more information from interacting with their environment than just the reward signal at the end of each episode.”

Source

Some thoughts on the Sutton interview— Dwarkesh Patel Podcast
Created: 6/13/2026, 3:28:01 AM

Your notes

Loading notes...