Training-free entropy-gated explicit<->latent (SwiReasoning) decoding for llama.cpp — bit-exact, Pareto-superior on a 27B
-
Updated
Jun 18, 2026 - C++
Training-free entropy-gated explicit<->latent (SwiReasoning) decoding for llama.cpp — bit-exact, Pareto-superior on a 27B
Add a description, image, and links to the soft-embeddings topic page so that developers can more easily learn about it.
To associate your repository with the soft-embeddings topic, visit your repo's landing page and select "manage topics."