Generative PCA replay: mean + 10 PCA directions per class hits 90.2%
BlackHC · x · 2026-09-16
The author's (generative) PCA replay reaches 90.2%: summarize each class's pixels with a mean plus 10 PCA directions, sample synthetic inputs, and save predictions before each new task while penalizing changes to them. The switch to summed all-class BCE means replay's effect isn't fully isolated.
Related event: BlackHC's toy continual learning experiment: PCA replay hits 91.1%(6 posts)→
More from Research
- After flybrain sim called 'just animations', researcher plans a verifiable simulation — TinfoilTricorn · 2026-09-16
- Well-funded AI4Science startups bet epistemology is optional for doing science — ludwigABAP · 2026-09-16
- An interactive Margolus-neighborhood cellular automaton where sand falls and water flows — mhmazur · 2026-09-16
- antirez open-sources DwarfStar's fused inference kernels under MIT — antirez · 2026-09-16
- The unreasonable effectiveness of BM25 for agentic search — Jo Kristian Bergum, Hornet.dev — AI Engineer · 2026-09-16
- Berkeley team shows probes can steer models to any level of risk aversion — soumitrashukla9 · 2026-09-16