Metraipolis › archive › dc_449e24e0bf89

TinyCeNN-LM: Quality-Gated Conversion of Pretrained Attention with CeNN-Inspired Cellular-Recurrent Layers — arXiv:2609.21139

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.21139
License (as archived)
CC0-1.0
Archived
2026-10-10T01:01:42.977Z
SHA-256 of the archived text
3466e8526dba2b4aaeed8754053082e2c5c7974f2ab78b6731530a55abeb3511
Stable URL of this page
https://metraipolis.com/documento/dc_449e24e0bf89

How to cite

TinyCeNN-LM: Quality-Gated Conversion of Pretrained Attention with CeNN-Inspired Cellular-Recurrent Layers — arXiv:2609.21139 — arXiv (en), https://arxiv.org/abs/2609.21139 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_449e24e0bf89 (Metraipolis archive dc_449e24e0bf89, sha256 3466e8526dba2b4aaeed8754053082e2c5c7974f2ab78b6731530a55abeb3511)
BibTeX
@misc{dc449e24e0bf89,
  title = {TinyCeNN-LM: Quality-Gated Conversion of Pretrained Attention with CeNN-Inspired Cellular-Recurrent Layers — arXiv:2609.21139},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.21139},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_449e24e0bf89; sha256 3466e8526dba2b4aaeed8754053082e2c5c7974f2ab78b6731530a55abeb3511},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21139. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Replacing attention in a pretrained language model is a compatibility problem: a plausible substitute may alter representations expected by later layers. TinyCeNN-LM introduces a \emph{quality-gated post-training conversion} framework using CeNN-inspired cellular-recurrent layers with bounded local processing, compact recurrent memory, routing, fusion, and accept-or-rollback validation. Three implementations are studied: Integrated Memory, MemoryFusion, and PDelta3-GDN2-CLVR+Local32. Strict PDelta3 conversion accepts a layer only when representation and NLL criteria pass fixed thresholds. On SmolLM2-135M, layers 0-2 are accepted with cumulative $\Delta\mathrm{NLL}=+0.01209$, while layer 3 is rejected despite acceptable NLL because representation fidelity fails. On Qwen3.5-0.8B, full-attention layers 3, 7, and 11 are accepted with final $\Delta\mathrm{NLL}=+0.02073$. Integrated Memory keeps perplexity within $-0.07\%$ to $+0.93\%$ while reducing total cache by up to $6.01\%$. A sampled 200-item downstream sanity check gives $28.5\%$--$32.0\%$ overall accuracy for converted Qwen releases. The results support conservative, quality-gated structural conversion rather than universal attention replacement or speedup.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.21139 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-10T00:23:42.430Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).