Author

Kazuhito Hotta

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Aug 2026

NeuroGuard: Neural Gradient Update Aware of Representation Damage

NeuroGuard is proposed, an update-control method added to DGR, a replay-based LT-CIL baseline, without adding learnable parameters, which preserves DGR's replay memory, classifier, and set of loss terms.

Taigo Sakai, Kazuhito Hotta · 0 citations