'Alien optimizers' rise: search and metalearning replace hand-designed learning rules
_rockt · x · 2026-11-12
Researchers coin "alien optimizers": learning rules found by search and metalearning rather than hand design. They argue the trend will accelerate, producing optimizers that co-evolve with architectures and are far more sample-efficient—at the cost of complexity. Experiments with the Eureka Machine (with Sakana AI collaborators) showcase the wild creativity of AI-discovered update rules.
More from Research
- Bronstein releases 78-page book: Mathematical Foundations of Geometric Deep Learning — caglar_ee · 2026-11-12
- Most biology AI models learn only one layer, and the big bet hinges on mapping — anshulkundaje · 2026-11-12
- Anthropic details agent memory: dreaming organizes Claude's memory, wiki+search hits 61% — RLanceMartin · 2026-11-12
- Talk preview: scaling 1,000 Fable 5.2 agents to formalize Fermat's Last Theorem — RLanceMartin · 2026-11-12
- Probing Qwen3 with a height ruler: each model's implicit "I" has a different height — Sauers_ · 2026-11-12
- Claude agent finds dwarf planet candidate in public sky data; $100K grants launched — jzlegion · 2026-11-12