A space of goals: the cognitive geometry of informationally bounded agents

Karen, Archer,; Volpi, Nicola Catenacci; Bröker, Franziska; Polani, Daniel

doi:10.48550/arxiv.2111.03699

Cited by 3 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are numerous analogies to be explored with respect to porting conceptual tools from relativity to study scale-free cognition. The use of cognitive geometry and infodesics [ 219 ] ties naturally to general relativity. Other examples include the following: Gravitational memory (permanent distortions of spacetime by gravitational waves [ 220 ]) to link the structure of action spaces to past experience; Inertia in terms of resilience to stress (anatomical homeostasis as a kind of inertia against movement in the morphospace and other spaces); Acceleration and force in a network space, where every connection in a network could be modeled via a “spring constant” or, even better, an LRC circuit.…”

Section: Implications: a Research Programmentioning

confidence: 99%

Competency in Navigating Arbitrary Spaces as an Invariant for Analyzing Cognition in Diverse Embodiments

Fields

Levin

2022

Entropy

113

View full text Add to dashboard Cite

One of the most salient features of life is its capacity to handle novelty and namely to thrive and adapt to new circumstances and changes in both the environment and internal components. An understanding of this capacity is central to several fields: the evolution of form and function, the design of effective strategies for biomedicine, and the creation of novel life forms via chimeric and bioengineering technologies. Here, we review instructive examples of living organisms solving diverse problems and propose competent navigation in arbitrary spaces as an invariant for thinking about the scaling of cognition during evolution. We argue that our innate capacity to recognize agency and intelligence in unfamiliar guises lags far behind our ability to detect it in familiar behavioral contexts. The multi-scale competency of life is essential to adaptive function, potentiating evolution and providing strategies for top-down control (not micromanagement) to address complex disease and injury. We propose an observer-focused viewpoint that is agnostic about scale and implementation, illustrating how evolution pivoted similar strategies to explore and exploit metabolic, transcriptional, morphological, and finally 3D motion spaces. By generalizing the concept of behavior, we gain novel perspectives on evolution, strategies for system-level biomedical interventions, and the construction of bioengineered intelligences. This framework is a first step toward relating to intelligence in highly unfamiliar embodiments, which will be essential for progress in artificial intelligence and regenerative medicine and for thriving in a world increasingly populated by synthetic, bio-robotic, and hybrid beings.

show abstract

Section: Implications: a Research Programmentioning

confidence: 99%

Competency in Navigating Arbitrary Spaces as an Invariant for Analyzing Cognition in Diverse Embodiments

Fields

Levin

2022

Entropy

113

View full text Add to dashboard Cite

show abstract

“…There has been a growing body of research aimed towards understanding, from a geometric perspective, how deep learning methods transform input data into decisions, memories, or actions [HR17, LAG + 20, SPG + 21]. Some studies have investigated how dimensionality reduction techniques can represent the geometry of task domains [SMK11] or the intrinsic geometry of a single agent's task [AVBP21]. However, such studies do not usually incorporate the geometry of the originating domain or task in a substantial way, before applying or investigating the performance of learning algorithms -and even fewer do so for multi-agent systems.…”

Section: Introductionmentioning

confidence: 99%

Detecting danger in gridworlds using Gromov's Link Condition

Burns¹,

Tang²

2022

Preprint

View full text Add to dashboard Cite

Gridworlds have been long-utilised in AI research, particularly in reinforcement learning, as they provide simple yet scalable models for many real-world applications such as robot navigation, emergent behaviour, and operations research. We initiate a study of gridworlds using the mathematical framework of reconfigurable systems and state complexes due to Abrams, Ghrist & Peterson. State complexes represent all possible configurations of a system as a single geometric space, thus making them conducive to study using geometric, topological, or combinatorial methods. The main contribution of this work is a modification to the original Abrams, Ghrist & Peterson setup which we believe is more naturally-suited to the context of gridworlds. With this modification, the state complexes may exhibit geometric defects (failure of Gromov's Link Condition), however, we argue that these failures can indicate undesirable or dangerous states in the gridworld. Our results provide a novel method for seeking guaranteed safety limitations in discrete task environments with single or multiple agents, and offer potentially useful geometric and topological information for incorporation in or analysis of machine learning systems.

show abstract

Expert Navigators Deploy Rational Complexity-Based Decision Prioritization for Large-Scale Real-World Planning

Fernandez Velasco,

Griesbauer,

Brunec

et al. 2024

Preprint

View full text Add to dashboard Cite

Efficient planning is a distinctive hallmark of human intelligence. Computational analyses of human behavior during planning provide evidence for hierarchically segmented representations of the state space of options. However, such evidence derives from simplistic tasks with small state-spaces that belie the complexity of real-world planning. Here, we examine the street-by-street route plans of London taxi drivers navigating across more than 26,000 streets in London (UK). Response times were faster for states with higher successor representations, providing evidence for predictive mapping. We also find an effect for the interaction between the successor representation of a state and local transition entropy, which indicates hierarchical chunking of transition sequences, and thus provides real-world support to existing theories of hierarchical state-space representations and planning. Finally, we explored how planning unfolded dynamically over different phases of the constructed journey and identify theoretic principles by which expert planners rationally prioritize specific states during the planning process. Overall, our findings provide real-world evidence for predictive maps and rational hierarchical prioritization in expert route planning.

show abstract

A space of goals: the cognitive geometry of informationally bounded agents

Cited by 3 publications

References 0 publications

Competency in Navigating Arbitrary Spaces as an Invariant for Analyzing Cognition in Diverse Embodiments

Competency in Navigating Arbitrary Spaces as an Invariant for Analyzing Cognition in Diverse Embodiments

Detecting danger in gridworlds using Gromov's Link Condition

Expert Navigators Deploy Rational Complexity-Based Decision Prioritization for Large-Scale Real-World Planning

Contact Info

Product

Resources

About