paper-with-me

Papers

Isotropy Cliffs: The Geometric Signature of Decision-Making in Large Language Models

2026-08-01 · Okan S. Coskun, Florian Rottach, Carsten Eickhoff, William Rudman arxiv

We investigate the geometry of decision-making in Multiple Choice Question Answering (MCQA) through the lens of isotropy. Analyzing five open-weight models across diverse datasets, we identify decision-critical transition layers characterized by a shift in isotropy, coinciding with a major representational change and the emergence of task-relevant clusters. We demonstrate that this synchronized geometric behavior is strongly correlated with downstream accuracy ($r\approx0.84$), displaying its relevance for successful decision-making. Furthermore, we show that this transition is robust to prompt variations, suggesting that it reflects a general mechanism of model behavior.

📄 PDF Abstract BibTeX arXiv:2608.00828

Code (0)

등록된 구현이 없습니다.

Tasks

Question Answering

Similar Papers 제목 키워드 기반

Anisotropic Random Feature Regression in High Dimensions

2021-09-29 · ICLR 2022 4 · Gabriel Mel, Jeffrey Pennington

In contrast to standard statistical wisdom, modern learning algorithms typically find their best performance in the overparameterized regime in which the model has many more parameters than needed to fit the training dat…

regressionVocal Bursts Intensity Prediction

Finding signatures of low-dimensional geometric landscapes in high-dimensional cell fate transitions

2025-06-04 · Maria Yampolskaya, Laertis Ikonomou, Pankaj Mehta

Multicellular organisms develop a wide variety of highly-specialized cell types. The consistency and robustness of developmental cell fate trajectories suggests that complex gene regulatory networks effectively act as lo…

Decision Making

How Does Fine-tuning Affect the Geometry of Embedding Space: A Case Study on Isotropy

2021-09-10 · Findings (EMNLP) 2021 11 · Sara Rajaee, Mohammad Taher Pilehvar

It is widely accepted that fine-tuning pre-trained language models usually brings about performance improvements in downstream tasks. However, there are limited studies on the reasons behind this effectiveness, particula…

Detecting LLM Hallucinations via Embedding Cluster Geometry: A Three-Type Taxonomy with Measurable Signatures

2026-02-15 · Matic Korun arxiv

We propose a geometric taxonomy of large language model hallucinations based on observable signatures in token embedding cluster structure. By analyzing the static embedding spaces of 11 transformer models spanning encod…

Signature-Informed Transformer for Asset Allocation

2025-10-03 · Yoontae Hwang, Stefan Zohren arxiv

Modern deep learning for asset allocation typically separates forecasting from optimization. We argue this creates a fundamental mismatch where minimizing prediction errors fails to yield robust portfolios. We propose th…

Decision Making