causal
Training on code improves general reasoning
Training models on coding makes them more rigorous and better at reasoning across many domains, even for users who never ask coding questions—so Meta heavily emphasized code in Llama-3 training despite WhatsApp users rarely coding.
causalpending
Speaker
Mark ZuckerbergEvidence Quote
“training the models on coding helps them become more rigorous in answering the question and helps them reason across a lot of different types of domains”
Source
Mark Zuckerberg — Llama 3, $10B models, Caesar Augustus, & 1 GW datacenters— Dwarkesh Patel PodcastCreated: 6/13/2026, 3:27:56 AM
Your notes
Loading notes...