Richard Sutton Publishes 16th PhD Student's Thesis on Preventing Plasticity Loss in Neural Networks
RichardSSutton · x · 2026-10-04
Richard Sutton announced that his 16th PhD student, Fernando Hernandez Garcia, has released his dissertation: Selective Reinitialization Algorithms for Preventing Plasticity Loss in Artificial Neural Networks.
- The thesis tackles continual learning — systems that must keep adapting to non-stationary data, a longstanding goal of AI.
- Its core focus is loss of plasticity: the phenomenon where neural networks progressively lose the ability to learn from new data.
- It proposes selective reinitialization algorithms to prevent this degradation and keep deep networks adaptable.
More from Research
- LEV: Looped decision model lifts 2.6B accuracy 63.2% to 76.2% with adaptive compute — jiqizhixin · 2026-11-19
- Aurora2-400M-A20M: a sparse MoE pretrained from scratch in 36 hours on one RTX Pro 6000 — Tall_Abrocoma_3533 · 2026-11-19
- Researcher calls METR/Anthropic '2x speedup' RSI analysis statistically dubious — jessi_cata · 2026-11-19
- Linear Probes Could Reveal Which Persona a Base Model Is Playing, Researchers Suggest — cephaloform · 2026-11-19
- ETH's Anymal robot completes 10km autonomous alpine hike with VLM-guided ROAM system — ChongZzZhang · 2026-11-19
- OpenVax codebase grows from 78K to 565K lines in 15 months — iskander · 2026-11-19