The paper examines how to allocate attention heads and head dimensions across Transformer layers to balance expressivity and efficiency. It provides a mathematical analysis of early layers’ role in information extraction and characterizes the trade‑off between head count and dimension under a fixed parameter budget. The authors prove a saturation effect of softmax activations, showing that increasing head dimensions yields diminishing returns, especially for long sequences, and propose strategies for efficient parameter allocation across layers.
arXiv:2607. 00479v1 Announce Type: new Abstract: Transformer-based large models have demonstrated remarkable generalization abilities across different tasks by leveraging a context-aware attention module for in-context learning.
arXiv:2608. 09558v1 Announce Type: new Abstract: How expressive is prompting a transformer?
arXiv:2610.07676v1 Announce Type: cross Abstract: Research on transformer expressivity shows whether a transformer is capable of solving a given task, but gives little indication of whether the solut...
arXiv:2609.01231v1 Announce Type: new Abstract: We prove that a multi-head scaled dot product attention can be viewed as a parameter identification strategy. The ratio of unidentified parameters to t...
arXiv:2610.26974v1 Announce Type: new Abstract: We show that two-layer transformers trained to near-zero loss on a sorting task converge to one of two mechanistically distinct solution classes depend...
arXiv:2601.11618v4 Announce Type: replace-cross Abstract: An operation's input-output function does not, by itself, determine its computational role within a represented process. That role also depen...
arXiv:2605. 18079v2 Announce Type: replace Abstract: Existing expressivity results for transformers typically rely on hardmax attention, high precision, and other architectural modifications that disconnect them from the models used in practice.
arXiv:2606. 27449v1 Announce Type: new Abstract: Multi-head attention conventionally partitions the hidden dimension equally across all heads at every layer, enforcing an identical representational subspace dimension (dh = dmodel/h) throughout the models depth.
arXiv:2602.06471v2 Announce Type: replace Abstract: The architectural shape of dense Transformers has remained remarkably stable: narrow-wide-narrow feed-forward networks (FFNs) consume most non-embe...
arXiv:2606. 29256v1 Announce Type: cross Abstract: In recent years, models based on the Transformer architecture have seen widespread applications and have become one of the core tools in the field of deep learning.
arXiv:2610.19358v1 Announce Type: new Abstract: In this paper we study the behavior of transformer networks generating text over long periods of time. We show that even a three-token, single-layer at...
The paper presents a retraining‑free technique that transforms pre‑trained transformers into elastic models capable of adapting to any computational budget on the fly. By jointly ranking every MLP neuron and attention head on a common scale and applying soft pruning across multiple sparsity targets, the method achieves smooth performance degradation up to 60% sparsity while outperforming one‑shot baselines. Experiments on 27 vision and text transformer encoders show significant parameter reductions with minimal accuracy loss and notable throughput and latency gains on GPUs, all produced in minutes.