Tersyn Labs

non-profit
Activity Feed

AI & ML interests

We are interested in small language models, model compression, and efficient neural architectures. Most of our work focuses on getting more capability out of fewer parameters through quantization, pruning, distillation, parameter sharing, and better training methods. We are also interested in multimodal models, mechanistic interpretability, and understanding which parts of a model actually matter for a given capability.

Recent Activity

sz14  updated a Space about 9 hours ago
Tersyn-Labs/README
sz14  published a Space about 9 hours ago
Tersyn-Labs/README
View all activity

Organization Card

Tersyn-Labs-Banner

models 0

None public yet

datasets 0

None public yet