LeCun vs. Dietterich: Should We Prefer Models Expressed by Short Formulas?
tdietterich · x · 2026-10-12
Yann LeCun questions whether preferring models expressible with a few short formulas—an inductive bias borrowed from symbolic regression—is actually justified. Thomas Dietterich responds that what matters isn't algebraic expressibility but systematic behavior across a wide range of conditions, which local kernel interpolation cannot provide.
Related event: LeCun and Dietterich Debate Whether LLMs Learn Laws or Just Interpolate(4 posts)→
More from Research
- Billion AI scientists will tackle aging, predicts AI commentator — Dr_Singularity · 2026-10-12
- RSIArena Stage 1: GPT-6 Astra wins blind vote as 10 AI agents self-train on Nemotron — my_cat_can_code · 2026-10-12
- Anshul Kundaje advises young engineer to present bio-lab replication results with more humility — anshulkundaje · 2026-10-12
- Gadsby-4B Hits Hugging Face: Constrained Decoding Turns Gemma 3 4B Into a Lipogram Model — cephaloform · 2026-10-12
- Weco AI Workshop Breaks Down AutoResearch Loop Behind OpenAI's Parameter Golf Challenge — AI Engineer · 2026-10-12
- Normal Computing ML head revamps site showcasing thermodynamic computing chip research — thomasahle · 2026-10-12