Metraipolis › archive › dc_488145c0efe0

Attention is Case-Sensitive — arXiv:2608.03711

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.03711
License (as archived)
CC0-1.0
Archived
2026-10-31T12:31:25.770Z
SHA-256 of the archived text
7319d35d02087afbd1d95a4c0a488d280606ff6bdfc6644a23ede4ba7f1f44f7
Stable URL of this page
https://metraipolis.com/documento/dc_488145c0efe0

How to cite

Attention is Case-Sensitive — arXiv:2608.03711 — arXiv (en), https://arxiv.org/abs/2608.03711 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_488145c0efe0 (Metraipolis archive dc_488145c0efe0, sha256 7319d35d02087afbd1d95a4c0a488d280606ff6bdfc6644a23ede4ba7f1f44f7)
BibTeX
@misc{dc488145c0efe0,
  title = {Attention is Case-Sensitive — arXiv:2608.03711},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.03711},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_488145c0efe0; sha256 7319d35d02087afbd1d95a4c0a488d280606ff6bdfc6644a23ede4ba7f1f44f7},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.03711. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In human visual perception, uppercase lettering serves as a natural salience cue that captures attention within lowercase text. In this paper, we present a systematic empirical characterization study revealing that Large Language Models (LLMs) exhibit an analogous property: letter casing modulates internal attention allocation. Through analysis across 13 models, nine LLMs and four Vision-Language Models (VLMs), with diverse tokenization schemes, we show that formatting target information in alternating or uppercase against a lowercase context concentrates attention on those textual spans. In text this effect is universal, holding across every evaluated non-reasoning model. We frame it as a previously under-explored latent property of pretrained transformers rather than a prescriptive method. Our investigation reveals a central attention-performance divergence: while this "casing effect" robustly shifts attention, its impact on downstream accuracy is non-trivial, increased concentration does not inherently improve task accuracy and, in high-entropy contexts like alternating case, can degrade it. We further identify a boundary condition: the deliberative "thinking" phase in reasoning models acts as a semantic buffer that mitigates typographic sensitivity in text. Extending the study to VLMs, we find the effect transfers partially: the same prompt-side casing reorganizes cross-modal attention along two coupled axes, predominantly a macroscopic disengagement from the image toward the text prompt, and secondarily a concentration of the residual visual attention on the target region. By isolating casing as a zero-shot mechanism for attention steering that requires no model access or fine-tuning, we provide a new foundational understanding of how pretraining internalizes typographic emphasis.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-05 in cs.CL).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Computation and Language (cs.CL); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.03711 — 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 2026-10-31T12:17:37.813Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-05?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).