English
Related papers

Related papers: Scale Determines Whether Language Models Organize …

200 papers

Understanding the latent space geometry of large language models (LLMs) is key to interpreting their behavior and improving alignment. Yet it remains unclear to what extent LLMs linearly organize representations related to semantic…

Computation and Language · Computer Science 2026-01-22 Baturay Saglam , Paul Kassianik , Blaine Nelson , Sajana Weerawardhena , Yaron Singer , Amin Karbasi

Geometric properties of Transformer weights, particularly the unembedding matrix, have been widely useful in language model interpretability research. Yet, their utility for estimating downstream performance remains unclear. In this work,…

Computation and Language · Computer Science 2026-02-25 Atharva Kulkarni , Jacob Mitchell Springer , Arjun Subramonian , Swabha Swayamdipta

Understanding what defines a good representation in large language models (LLMs) is fundamental to both theoretical understanding and practical applications. In this paper, we investigate the quality of intermediate representations in…

Machine Learning · Computer Science 2024-12-13 Oscar Skean , Md Rifat Arefin , Yann LeCun , Ravid Shwartz-Ziv

Why do language models from different architecture families respond so differently to the same perturbation? We argue that the answer is not scale, but \emph{how architecture shapes information compression}. Analyzing eight Transformer…

Computation and Language · Computer Science 2026-05-07 Yukin Zhang , Qi Dong , Kemu Xu

How do transformer language models represent magnitude? Recent work disagrees: some find logarithmic spacing, others linear encoding, others per-digit circular representations. We apply the formal tools of psychophysics to resolve this.…

Computation and Language · Computer Science 2026-03-24 Jon-Paul Cacioli

Large Language Models (LLMs) show strong generalization across diverse tasks, yet the internal decision-making processes behind their predictions remain opaque. In this work, we study the geometry of hidden representations in LLMs through…

Machine Learning · Computer Science 2025-11-26 Abhinav Joshi , Divyanshu Bhatt , Ashutosh Modi

Recent work has shown that small transformers trained in controlled "wind-tunnel'' settings can implement exact Bayesian inference, and that their training dynamics produce a geometric substrate -- low-dimensional value manifolds and…

Machine Learning · Computer Science 2026-05-19 Naman Agarwal , Siddhartha R. Dalal , Vishal Misra

We present Multi-Scale Manifold Alignment(MSMA), an information-geometric framework that decomposes LLM representations into local, intermediate, and global manifolds and learns cross-scale mappings that preserve geometry and information.…

Computation and Language · Computer Science 2025-10-14 Yukun Zhang , Qi Dong

We show that a language model's ability to predict text is tightly linked to the breadth of its embedding space: models that spread their contextual representations more widely tend to achieve lower perplexity. Concretely, we find that…

Computation and Language · Computer Science 2026-04-21 Yanhong Li , Ming Li , Karen Livescu , Jiawei Zhou

What does a world model learn from physical exploration, without any linguistic supervision? We argue the answer is organized by a single principle: the geometric structure of the physical world. Training a VAE-based world model on random…

Machine Learning · Computer Science 2026-05-29 Jiayi Fang

Large language models (LLMs) work by manipulating the geometry of input embedding vectors over multiple layers. Here, we ask: how are the input vocabulary representations of language models structured, and how and when does this structure…

Computation and Language · Computer Science 2025-10-10 Isabel Papadimitriou , Jacob Prince

From extracting features to generating text, the outputs of large language models (LLMs) typically rely on the final layers, following the conventional wisdom that earlier layers capture only low-level cues. However, our analysis shows that…

Machine Learning · Computer Science 2025-06-17 Oscar Skean , Md Rifat Arefin , Dan Zhao , Niket Patel , Jalal Naghiyev , Yann LeCun , Ravid Shwartz-Ziv

Understanding the representations of different languages in multilingual language models is essential for comprehending their cross-lingual properties, predicting their performance on downstream tasks, and identifying any biases across…

Computation and Language · Computer Science 2023-05-16 Cheril Shah , Yashashree Chandak , Manan Suri

We investigate the geometry of predictive information across the layers of large language models (LLMs). We repurpose representation lenses-learned affine maps trained to predict the next token from intermediate residual streams-as…

Machine Learning · Computer Science 2026-05-12 Gianfranco Lombardo , Giuseppe Trimigno , Stefano Cagnoni

The internal representations learned by language models consistently exhibit striking geometric structure: calendar months organize into a circle, historical years form a smooth one-dimensional manifold, and cities' latitudes and longitudes…

Machine Learning · Computer Science 2026-02-27 Dhruva Karkada , Daniel J. Korchinski , Andres Nava , Matthieu Wyart , Yasaman Bahri

A central question in cognitive science is whether conceptual representations converge onto a shared manifold to support generalization, or diverge into orthogonal subspaces to minimize task interference. While prior work has discovered…

Computation and Language · Computer Science 2026-02-09 Zhimin Hu , Lanhao Niu , Sashank Varma

How large language models (LLMs) align with the neural representation and computation of human language is a central question in cognitive science. Using representational geometry as a mechanistic lens, we addressed this by tracking…

Neurons and Cognition · Quantitative Biology 2026-02-10 Yixuan Liu , Zhiyuan Ma , Likai Tang , Runmin Gan , Xinche Zhang , Jinhao Li , Chao Xie , Sen Song

Next-token predictors often appear to develop internal representations of the latent world and its rules. The probabilistic nature of these models suggests a deep connection between the structure of the world and the geometry of probability…

Machine Learning · Computer Science 2026-03-18 Sasha Brenner , Thomas R. Knösche , Nico Scherf

In cognitive science and AI, a longstanding question is whether machines learn representations that align with those of the human mind. While current models show promise, it remains an open question whether this alignment is superficial or…

Neurons and Cognition · Quantitative Biology 2025-10-27 Craig Sanders , Billy Dickson , Sahaj Singh Maini , Robert Nosofsky , Zoran Tiganj

Vision-language models encode continuous geometry that their text pathway fails to express: a 6,000-parameter linear probe extracts hand joint angles at 6.1 degrees MAE from frozen features, while the best text output achieves only 20.0…

Computer Vision and Pattern Recognition · Computer Science 2026-03-09 Yakov Pyotr Shkolnikov
‹ Prev 1 2 3 10 Next ›