Metraipolis › archive › dc_ec71b01d8f72

TopoQuant: Topological Framework to Characterize Behavioral Shifts in Quantized Large Language Models — arXiv:2611.00631

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00631
License (as archived)
CC0-1.0
Archived
2026-11-03T11:01:25.167Z
SHA-256 of the archived text
0af757657d9cc4167954ccc681cd800197baf468519ea1921b637c71f3fd6279
Stable URL of this page
https://metraipolis.com/documento/dc_ec71b01d8f72

How to cite

TopoQuant: Topological Framework to Characterize Behavioral Shifts in Quantized Large Language Models — arXiv:2611.00631 — arXiv (en), https://arxiv.org/abs/2611.00631 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ec71b01d8f72 (Metraipolis archive dc_ec71b01d8f72, sha256 0af757657d9cc4167954ccc681cd800197baf468519ea1921b637c71f3fd6279)
BibTeX
@misc{dcec71b01d8f72,
  title = {TopoQuant: Topological Framework to Characterize Behavioral Shifts in Quantized Large Language Models — arXiv:2611.00631},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00631},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ec71b01d8f72; sha256 0af757657d9cc4167954ccc681cd800197baf468519ea1921b637c71f3fd6279},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00631. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Post-training quantization (PTQ) reduces the com putational cost of large language models (LLMs), but it can introduce internal structural and behavioral changes that are not captured by traditional metrics such as accuracy and per plexity. We present TopoQuant, a topological framework that uses topological data analysis (TDA) to quantify structural drift in self-attention weights and relate it to behavioral changes in quantized LLMs. Across four LLM families and multiple quantization levels, we show that topological drift remains stable at moderate compression levels but increases sharply below 4 bit precision, closely tracking declines in attention agreement. Our results reveal a strong relationship between topological drift and attention agreement, indicating that structural changes in model parameters closely reflect behavioral degradation under quantization. We further demonstrate that topological features can accurately predict attention consistency using lightweight machine learning models, achieving R2 scores of up to 0.85. Predicting attention agreement enables efficient assessment of be havioral consistency in quantized LLMs, allowing quantization induced degradation to be detected without running expensive downstream tasks. The results indicate that topological drift pro vides an efficient early-warning signal for quantization-induced degradation and can help identify safe compression limits without extensive downstream evaluation.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.00631 — 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 2026-11-03T10:59:45.732Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).