François Fleuret explains a memory trick: pay O(time steps) instead of O(steps × components)
francoisfleuret · x · 2026-11-17
Former Meta/Apple research scientist François Fleuret shares what he calls a "wonderful trick" for state memory: instead of storing the state as S(t)={x1(t),...,xt(t)} (each step holding only current components), keep the full history S(t)={x1(1),...,xt(t)}, cutting memory from O(nbtimesteps × nbstatecomponents) to O(nbtimesteps) — while you want both factors large. Useful for long-sequence / long-horizon state modeling.
More from Research
- 8th NeurIPS Robot Learning Workshop to debate: Is Physical AI truly going zero-shot? — m_wulfmeier · 2026-11-17
- François Fleuret dissects a fundamental transformer limitation: sequence and token indexing — francoisfleuret · 2026-11-17
- Fudan & OpenMOSS map the trade-off: long-context hybrids learn more, extrapolate less reliably — jiqizhixin · 2026-11-17
- Models favor their own APIs in code: Opus calls Claude API in 97% of implementations — santiviquez · 2026-11-17
- Search behavior gap: Opus searched in 10/60 answers, GPT models searched every time — santiviquez · 2026-11-17
- 79% of 2,880 LLM API recommendations went to just four big-cloud services — santiviquez · 2026-11-17