Update README.md
Browse files
README.md
CHANGED
|
@@ -68,6 +68,15 @@ If you use or build upon this demo, please cite:
|
|
| 68 |
url={https://arxiv.org/abs/2507.04886},
|
| 69 |
}
|
| 70 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
```
|
| 72 |
|
| 73 |
This work demonstrates that transformer blocks, not token embeddings, carry the semantic burden in LLMs — a step toward modular, fusable, multilingual LMs.
|
|
|
|
| 68 |
url={https://arxiv.org/abs/2507.04886},
|
| 69 |
}
|
| 70 |
|
| 71 |
+
@misc{bochkov2025growingtransformersmodularcomposition,
|
| 72 |
+
title={Growing Transformers: Modular Composition and Layer-wise Expansion on a Frozen Substrate},
|
| 73 |
+
author={A. Bochkov},
|
| 74 |
+
year={2025},
|
| 75 |
+
eprint={2507.07129},
|
| 76 |
+
archivePrefix={arXiv},
|
| 77 |
+
primaryClass={cs.LG},
|
| 78 |
+
url={https://arxiv.org/abs/2507.07129},
|
| 79 |
+
}
|
| 80 |
```
|
| 81 |
|
| 82 |
This work demonstrates that transformer blocks, not token embeddings, carry the semantic burden in LLMs — a step toward modular, fusable, multilingual LMs.
|