A Logic of General Attention Using Edge-Conditioned Event Models

Authors: Gaia Belardinelli, Thomas Bolander, Sebastian Watzl

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this work, we present the first general logic of attention. Attention is a powerful cognitive ability that allows agents to focus on potentially complex information, such as logically structured propositions, higher-order beliefs, or what other agents pay attention to. ... We provide the first principles for general attention that can be used to axiomatize specific attention notions.
Researcher Affiliation Academia Gaia Belardinelli1 , Thomas Bolander2 , Sebastian Watzl3 1Stanford University 2Technical University of Denmark 3University of Oslo EMAIL, EMAIL, EMAIL
Pseudocode No The paper describes logical definitions, theorems, and proofs but does not contain any sections explicitly labeled "Pseudocode" or "Algorithm", nor does it present structured steps in a code-like format.
Open Source Code No Due to page limits, we only provide proof sketches. All full proofs can be found in the extended versions on Arxiv: https://arxiv.org/pdf/2505.14539. This link provides extended proofs, not source code for the methodology.
Open Datasets No The paper is theoretical, introducing a logical framework for attention. It uses illustrative examples (e.g., "Ann is about to review the CV of an applicant") rather than empirical datasets. No specific dataset, link, DOI, or citation to a publicly available dataset is provided.
Dataset Splits No The paper does not use any datasets for empirical experiments, therefore no information regarding training/test/validation splits is provided.
Hardware Specification No This paper presents a theoretical framework and does not report on empirical experiments that would require specific hardware. Therefore, no hardware specifications are mentioned.
Software Dependencies No The paper introduces a logical system and does not describe any implementation or experimental results that would necessitate listing specific software dependencies with version numbers.
Experiment Setup No As a theoretical paper, it does not involve empirical experiments with specific setup details such as hyperparameters, training configurations, or system-level settings.