Framework clarifies what acquired maps reveal about cognitive skills
Map Users and Mapmakers: The Scope of Cognitive Attribution from Acquired Representations
Artificial Intelligence
Summary
This paper explores how using maps or acquired mental pictures can increase what a system understands or can do. The authors show that having a map doesn’t always mean you know how the map was made or can figure out new things from scratch. They create a table that breaks down these different abilities to better understand what is really learned and what remains unknown. The work helps explain why recognizing or using concepts is not the same as discovering or fully understanding them. They test this idea with examples like game-playing models and machine learning concepts.
What this means in practice
- •For ai developers: Design training pipelines that distinguish between competence application and concept acquisition to better interpret AI model capabilities.
- •For cognitive system designers: Develop systems with clear attribution of capacities, avoiding overestimating what concepts the system truly understands versus what it only applies.
A position paper. It proposes an approach and reports no results.
Authors
Yiling Wu
Abstract
An acquired representation can enlarge a system's cognitive repertoire without transferring the capacities exercised in producing that representation. This paper develops a framework for specifying that enlargement and its limits. Its central contribution is a five-part attribution table distinguishing effective tracking, application of acquired structures, acquisition from explicit specifications, acquisition from identifying observations, and retention and reuse. Each entry identifies a positive capacity commitment and a further claim requiring additional support. The argument deliberately grants meaningful content, causal efficacy, and productive inference, so that its conclusion does not depend on treating representations as inert encodings. Map and category examples show why even complete application competence leaves acquisition capacity undetermined, and why acquiring a criterion from its description differs from finding it in examples. Short formal proofs appear in an appendix. The framework is applied to Andrew Ng's world-model interpretation of Othello-GPT and to the specific indicators discussed in contemporary accounts of machine concepts. It preserves demonstrated recognition, classification, inference, and qualified acquisition while specifying what remains unestablished about criterion discovery and accumulation. The result concerns the scope of cognitive attribution rather than the constitutive conditions of concept possession: cognitive achievements deserve credit for the capacities they establish, without silently importing a broader repertoire through the labels attached to them.