cd/sources/thomasdullien-auto-discoveredยท homeโ€บ sourcesโ€บ Thomasdullien (auto-discovered)
cat /sources/thomasdullien-auto-discovered.feed | wc -l โ†’ 1

Thomasdullien (auto-discovered)

articles 1 domain thomasdullien.github.io โ†’ feed RSS
03:48
2026-07-19
thomasdullien.github.io
large-language-models

RL economics, morally charged terms, and "distillation"

Reinforcement learning (RL) is the primary driver of recent advances in coding and mathematics for large language models (LLMs), according to an analysis of the economics of model improvement after huโ€ฆ