English
Related papers

Related papers: Relational Rank Geometry in Transformers: Detectin…

200 papers

Large reasoning models (LRMs) enhance problem-solving capabilities by generating explicit multi-step chains of thought (CoT) reasoning; however, they incur substantial inference latency and computational overhead. To mitigate this issue,…

Artificial Intelligence · Computer Science 2026-04-21 Jiayi Tian , Yupeng Su , Ryan Solgi , Souvik Kundu , Zheng Zhang

Selecting the latent dimensions (ranks) in tensor factorization is a central challenge that often relies on heuristic methods. This paper introduces a rigorous approach to determine rank identifiability in probabilistic tensor models, based…

Machine Learning · Computer Science 2026-04-03 Eliezer da Silva , Arto Klami , Diego Mesquita , Iñigo Urteaga

Transformers face quadratic complexity and memory issues with long sequences, prompting the adoption of linear attention mechanisms using fixed-size hidden states. However, linear models often suffer from limited recall performance, leading…

Computation and Language · Computer Science 2025-07-10 Dustin Wang , Rui-Jie Zhu , Steven Abreu , Yong Shan , Taylor Kergan , Yuqi Pan , Yuhong Chou , Zheng Li , Ge Zhang , Wenhao Huang , Jason Eshraghian

Vision-language models (VLMs) are increasingly used to answer questions about physical scenes, yet most evaluations reduce performance to a final answer. This hides whether the model perceived the right objects, represented the right…

Computation and Language · Computer Science 2026-05-29 Emmanuelle Bourigault

Scientific time series often encode predictive geometric structure, including connectivity, cycles, shell-like geometry, directional changes, and nonlinear neighborhoods, that standard dot-product attention does not explicitly represent. We…

Machine Learning · Computer Science 2026-05-06 Usef Faghihi , Amir Saki

The topology of many real complex networks has been conjectured to be embedded in hidden metric spaces, where distances between nodes encode their likelihood of being connected. Besides of providing a natural geometrical interpretation of…

Physics and Society · Physics 2017-01-23 Antoine Allard , M. Ángeles Serrano , Guillermo García-Pérez , Marián Boguñá

Network embedding has attracted an increasing attention over the past few years. As an effective approach to solve graph mining problems, network embedding aims to learn a low-dimensional feature vector representation for each node of a…

Social and Information Networks · Computer Science 2020-08-10 Xiao Shen , Fu-Lai Chung

We investigate the relationship between the geometry of token embeddings and their role in the next token prediction within transformer models. An important aspect of this connection uses the notion of empirical measure, which encodes the…

Computation and Language · Computer Science 2025-01-22 Karthik Viswanathan , Yuri Gardinazzi , Giada Panerai , Alberto Cazzaniga , Matteo Biagetti

Transformer language models (LMs) have been shown to represent concepts as directions in the latent space of hidden activations. However, for any human-interpretable concept, how can we find its direction in the latent space? We present a…

Computation and Language · Computer Science 2024-04-02 David Chanin , Anthony Hunter , Oana-Maria Camburu

Estimating the difficulty of input questions as perceived by large language models (LLMs) is essential for accurate performance evaluation and adaptive inference. Existing methods typically rely on repeated response sampling, auxiliary…

Computation and Language · Computer Science 2025-09-17 Yubo Zhu , Dongrui Liu , Zecheng Lin , Wei Tong , Sheng Zhong , Jing Shao

Spatial reasoning requires both location-bound computation and location-invariant structure: agents must make local moves while preserving route, object, or constraint-level plans. We propose interaction locality, a task-geometry-aware…

Artificial Intelligence · Computer Science 2026-05-21 Yosuke Miyanishi , Tetsuro Morimura

A method is presented for accelerating inference in transformer language models by exploiting the low effective rank of the token activation manifold at each layer. The method decomposes each activation vector into a subspace component and…

Machine Learning · Computer Science 2026-05-06 Stephen J. Thomas

Current alignment evaluation mostly measures whether models encode dangerous concepts and whether they refuse harmful requests. Both miss the layer where alignment often operates: routing from concept detection to behavioral policy. We…

Machine Learning · Computer Science 2026-05-04 Gregory N. Frank

Relational concepts are indeed foundational to the structure of knowledge representation, as they facilitate the association between various entity concepts, allowing us to express and comprehend complex world knowledge. By expressing…

Computation and Language · Computer Science 2024-06-21 Zijian Wang , Britney White , Chang Xu

Given a $3$-manifold $M$ with multiple incompressible torus boundary components, we develop a general definition of order-detection of tuples of slopes on the boundary components of $M$. In parallel, we arrive at a general definition of…

Geometric Topology · Mathematics 2024-02-26 Adam Clay , Junyu Lu

Language recognition tasks are fundamental in natural language processing (NLP) and have been widely used to benchmark the performance of large language models (LLMs). These tasks also play a crucial role in explaining the working…

Machine Learning · Computer Science 2025-05-30 Ruiquan Huang , Yingbin Liang , Jing Yang

Identifying objects in an image and their mutual relationships as a scene graph leads to a deep understanding of image content. Despite the recent advancement in deep learning, the detection and labeling of visual object relationships…

Computer Vision and Pattern Recognition · Computer Science 2021-07-13 Rajat Koner , Poulami Sinhamahapatra , Volker Tresp

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. This work extends that foundational theory to the diverse…

Machine Learning · Computer Science 2025-12-24 Adrian Goldwaser , Michael Munn , Javier Gonzalvo , Benoit Dherin

Architectural obfuscation - e.g., permuting hidden-state tensors, linearly transforming embedding tables, or remapping tokens - has recently gained traction as a lightweight substitute for heavyweight cryptography in privacy-preserving…

Cryptography and Security · Computer Science 2025-06-24 Marcos Florencio , Thomas Barton

Residual mappings have been shown to perform representation learning in the first layers and iterative feature refinement in higher layers. This interplay, combined with their stabilizing effect on the gradient norms, enables them to train…

Machine Learning · Computer Science 2022-06-06 Mathias Lechner , Ramin Hasani , Zahra Babaiee , Radu Grosu , Daniela Rus , Thomas A. Henzinger , Sepp Hochreiter