Bochkov commited on
Commit
2205fdc
·
verified ·
1 Parent(s): cb2b9cf

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -0
README.md CHANGED
@@ -24,6 +24,8 @@ This repository provides **Model_64_FLOAT (272M)** — an **ablation model** fro
24
 
25
  [📚 Paper (Growing Transformers: Modular Composition and Layer-wise Expansion on a Frozen Substrate)](https://huggingface.co/papers/2507.07129) -
26
 
 
 
27
  This checkpoint tests whether language modeling and semantic structure can emerge when the **entire input embedding layer is frozen** and contains **no semantic or glyph/visual information**.
28
 
29
  Compared to **Model_64_BIT**, this model uses the same embedding dimensionality (`n_embed=64`) and the same “unique per token” construction, but the embedding vectors are **floating-point** (after a deterministic projection/normalization step), rather than raw binary components.
 
24
 
25
  [📚 Paper (Growing Transformers: Modular Composition and Layer-wise Expansion on a Frozen Substrate)](https://huggingface.co/papers/2507.07129) -
26
 
27
+ [📚 Blog Article](https://huggingface.co/blog/Bochkov/emergent-semantics-beyond-token-embeddings)
28
+
29
  This checkpoint tests whether language modeling and semantic structure can emerge when the **entire input embedding layer is frozen** and contains **no semantic or glyph/visual information**.
30
 
31
  Compared to **Model_64_BIT**, this model uses the same embedding dimensionality (`n_embed=64`) and the same “unique per token” construction, but the embedding vectors are **floating-point** (after a deterministic projection/normalization step), rather than raw binary components.