causal
Scaling laws are multiplicative
Because pre-training, post-training reinforcement learning, and test-time compute scaling laws multiply together, applying the new RL and test-time scaling to better base models compounds gains.
causalpending
Speaker
Gavin BakerEvidence Quote
“all these scaling laws are multiplicative. So, now we're going to apply these two new reinforcement learnings”
Created: 11/20/2026, 2:50:27 AM
Your notes
Loading notes...