BLAST RADIUS

Open-weight & owned stack · Training tuning

Lacuna Introduces Intermediate-Probe Knowledge Distillation

Nov 2, 2026

Lacuna introduced PROBE-KD, a distillation method that uses intermediate language-model representations to produce task-specific models.

Read the original at lacuna.tiptreesystems.comOpens the publisher's site in a new tab

More in Open-weight & owned stack

Meituan Releases LongCat-2.5 Multimodal ModelNov 2Dejan reproduces latent skills in GemmaNov 2Reflection AI Releases Beam Open-Weight ModelNov 2Google Meta Open-Source Native PyTorch TPU BackendNov 2DeepSeek adds Creator Mode for custom pluginsNov 2