arXiv:2606. 03398v1 Announce Type: cross Abstract: Formal languages have proven to be effective conduits to understand the inner mechanisms of transformers.
arXiv:2604. 22128v2 Announce Type: replace-cross Abstract: When trained on tasks requiring an understanding of hierarchical structure, transformers have been found to represent this hierarchy in distinct ways: in the geometry of the residual stream, and in stack-like attention patterns maintaining a last-in, first-out ordering.
The paper critiques the common practice of evaluating depth usage in depth‑recurrent language models by truncating depth during inference and measuring performance decline. It argues that this method conflates three distinct effects—fewer block applications, reduced computation, and an out‑of‑distribution readout—yet is usually interpreted as measuring only the second. To address this, the authors introduce the Depth Control Protocol (DCP), a suite of positive and negative controls that isolate each factor, along with a training intervention to confirm causality, specifically tailored for depth‑wise weight‑sharing architectures.
arXiv:2607. 01218v1 Announce Type: cross Abstract: Transformers use the same forward computation stream to both predict the next token and store useful state for future token predictions.
arXiv:2511. 17864v3 Announce Type: replace Abstract: Recent research has established that the impact of context in a vanilla transformer can be represented implicitly by forming a token-dependent, rank-1 patch to its MLP weights.
arXiv:2609.07876v1 Announce Type: cross Abstract: Recent methods in language model interpretability employ techniques such as sparse autoencoders to decompose residual stream contributions into linea...
The paper demonstrates that Large Language Models, despite their non‑linear components, exhibit a fundamental linearity property: when inputs from two distinct text streams are linearly combined, the model outputs a superposition of the individual next‑token distributions. This "Superposition Linearity Hypothesis" appears to be an intrinsic feature of the Transformer architecture, tends to weaken during pretraining, but can be largely restored with lightweight fine‑tuning. The authors also present a guided decoding method that separates the superposed outputs, allowing two coherent continuations to be generated from a single forward pass.
arXiv:2607. 18570v1 Announce Type: cross Abstract: Discourse relations provide document structure, critical to language understanding and enabling language model performance and ethicality.
The paper investigates how transformer models learn latent structure by training a small decoder-only transformer on three variants of the Alchemy benchmark. It finds that the model acquires different components of latent structure in discrete stages, with a notable asymmetry: it robustly composes fundamental transitions but struggles to decompose complex examples into atomic transitions. Layer‑specific causal interventions reveal plasticity windows where freezing layers delays or prevents stage completion, offering a detailed view of capability evolution during training.
Transformer language models process sequences token by token in an autoregressive manner, making growing contexts increasingly expensive. Yet many adjacent token spans are highly predictable or freque...
The paper investigates how large language models encode and use relational information among tokens across transformer layers. By analyzing activations from prompts that require inferring relationships among three cyclic tokens (months, hours, weekdays, musical notes), the authors find a consistent layerwise progression: intermediate layers capture pairwise relationships, while later layers encode the full three‑token relationship to predict the next token. They also identify geometrically structured token relationships that do not influence prediction, and show that constraining models to use only causally relevant joint representations improves next‑token accuracy.
We present a theoretical framework to explain the emergence of inductive reasoning abilities in Transformer language models. While previous works on Transformer learning dynamics have so far been mostly tied to specific tasks, we study a generalized class of inductive tasks that unifies several synthetic tasks known in the literature, including in-context n-grams and multi-hop reasoning.
Language modeling using Transformers has become commonplace despite their fixed computational depth and quadratic runtime with respect to input tokens. Recurrent models on the other hand offer linear...