Hacker GitHub

← All repos

ferreirafabio/autoresearch-automl

View on GitHub ↗ ★ 62 Python

Can LLMs beat classical HPO? A benchmark comparing classical, LLM-based, and hybrid methods on Karpathy's autoresearch.

3 mentions on Hacker News by 3 people: 1 story, 2 comments · first 2026-03-23 · latest 2026-06-09

Mentions per year 3 mentions · busiest year 2026 (3)
20082026
AllStoriesComments NewestTopOldest
  1. …easy to implement using a coding agent for the LLM and the ask-and-tell interface in pycma[2]. [1]: https://github.com/ferreirafabio/autoresearch-automl/blob/ma... [2]: https://github.com/CMA-ES/pycma
    comment by woadwarrior01 · 2026-06-09 · in Can LLMs Beat Classical Hyperparameter Optimization Algorithms?
  2. …context, one experiment showing that classical black-box optimization did better than Autoresearch: https://github.com/ferreirafabio/autoresearch-automl
    comment by strangecasts · 2026-03-23 · in 'The Karpathy Loop': 700 experiments, 2 days