arXiv:2608. 14689v1 Announce Type: new Abstract: Residual connections are a fundamental component of transformer architectures, yet the roles of the attention and feed-forward residual pathways remain poorly understood when considered independently.
arXiv:2607. 18363v1 Announce Type: cross Abstract: Feed-forward networks hold two thirds of a transformer's non-embedding parameters, yet the architecture has not received a necessity test that controls parameters, compute, and depth at once.
arXiv:2606. 31859v1 Announce Type: new Abstract: Residual connections add every sublayer's proposed update with a fixed coefficient of one; the network never evaluates whether an update is reliable before committing it.
arXiv:2607. 01940v1 Announce Type: cross Abstract: Mechanistic interpretability often relies on component-level interventions to discover how a model produces a behavior.
arXiv:2510. 04212v4 Announce Type: replace-cross Abstract: The pursuit of computational efficiency has driven the adoption of low-precision formats for training transformer models.
arXiv:2608. 02071v1 Announce Type: new Abstract: Attention-only dynamical theories model Transformer residual directions as particles aggregating on a sphere.
arXiv:2601. 22580v2 Announce Type: replace-cross Abstract: The success of Large Language Models (LLMs) hinges on the stable training of deep Transformer architectures.
arXiv:2603. 22473v2 Announce Type: replace-cross Abstract: Hybrid language models combine softmax attention with linear-time sequence mechanisms such as state-space or linear-attention layers, but the functional contribution of each component type remains insufficiently characterized.
arXiv:2606. 02559v1 Announce Type: cross Abstract: Post-training compression of Large Language Models (LLMs) removes entire architectural components, either deleting them or replacing them with fitted modules.
arXiv:2606. 01495v1 Announce Type: new Abstract: We present CART (Context-Anchored Recurrent Transformer), a parameter-efficient language model that reuses a single shared core block R times across depth.
arXiv:2606. 22248v2 Announce Type: replace Abstract: Standard autoregressive Transformer decoders can often exhibit substantial forgetting under sequential fine-tuning on shifting curriculum distributions.
arXiv:2608. 07921v1 Announce Type: cross Abstract: We apply Marchenko-Pastur (MP) random matrix theory to pre-trained attention weights in order to separate each projection matrix into a random-like bulk and a set of spectral outliers.
arXiv:2410. 24050v3 Announce Type: replace Abstract: Large-scale pretraining of transformers has been central to the success of foundation models.