It’s time for Embodied Intelligence 2024! Once again (March 20–22), we will be participating in a celebration of Computational Modeling, Robotics, CogSci, Neuroscience, Philosophy, and more! Recording of main session available on their YouTube channel.
The one presentation contributed by the lab is follow-up work on our Developmental Braitenberg Vehicles work (Representational Brain and Phenotypes group).
Developmental Neurosimulation as a Route to Embodied Morphogenetic-inspired Intelligence (Self-Organized Systems track)
The Developmental Neurosimulation approach allows us to approximate simple connectomes and their associated behaviors that emerge in an embodied context. By combining morphological and behavioral bio-inspiration with topological constructivism, the relationship between morphology and connectome have been demonstrated in three ways: providing a substrate for information acquisition, the role of pattern formation and developmental transformations in information processing, and the use of epigenetic functions to simulate critical periods and other forms of nonlinear acquisition. Embodied connectomes are constructed from innate instruction sets, which result in morphogenetic-inspired substrates for information acquisition. In particular, connecting a connectionist network with an embodied input/output network can act to structure input data with respect to orientation behaviors towards a stimulus. This results in a highly structured connectome topology exhibiting developmental freedom, or alternate forms of equivalent connectivity given plasticity and constraint. To grow topologically complex embodied connectionist networks, we utilize generative pattern formation as a mechanism for morphological transformation. This produces bodies of different shapes and internal networks of different sizes, and leads to acquisition that is grounded in how the agent actually utilizes the environment. Finally, successful developmental neurosimulation requires a variable model of temporal information processing that more closely resembles naturalistic interactions. To capture this in a computational model, we propose using epigenetic functions to enable stage-specific acquisition. Stage-specific acquisition, consistent with morphological development and enabling critical periods, structures association-building through increasing network capacity. Ultimately this contributes to an agent-based form of generative intelligence with a capacity for behavioral complexity. We will also demonstrate developmental neurosimulation with two examples: a sensory array influenced during a critical period, and the action of developmental freedom on a small, dense Braitenberg Vehicle connectome.
For more detail on some of the concepts discussed within, please check out these resources.
Continual Developmental Neurosimulation Using Embodied Computational Agents. arXiv, 2103.05753.
Critical Periods and Developmental Neurosimulation. PsyArXiv, doi:10. 31234/osf.io/jsqcy.
Jesse Parent and Hussain Ather also proposed talks for the breakout session track but were not able to present due to scheduling difficulties.
Jesse Parent
“The Mind as a ‘___’ ”: Past, Present, and Futures of Analogies of Mind & Brain and its Relationship to Technological Development
Comparisons of the “mind” (or the seat of mental or cognitive activities) to other, perhaps less esoteric objects, is a feature of interest across cultures and across periods of history. In this paper, we investigate the development of these analogies, as well as how they relate to different periods’ interpretations of how the mind works. We also consider how those discussions relate to the level of technology available to cultures or populations, and potentially associated views of self and others; the role of complex constructions or artifacts, and how this is related to nuance of what generates a person’s experience. Following broad historical analysis, we anchor this discussion in one of the most recent dominant analogies which has influenced popular discourse as well as afforded an attempted cross-disciplinary unification of scholarly activity — the academic arena of Cognitive Science and the likeness of a mind to a computational machine. We conclude with reflection on the nature of technological advancement — for mind & brain analogies, and sense-making in general — alongside technology’s role as a medium for inquiry, reference, and as a portal for understanding others.
Hussain Ather
Navigating Reality: Exploring Spatial Cognition, Virtual Environments, and Philosophical Implications (Morphological Computation track)
Spatial navigation research, rooted in cognitive science, offers insights into how humans and animals process spatial information, shaping cognitive maps of their environments. These inquiries extend beyond understanding cognitive mechanisms, delving into philosophical debates on realism and anti-realism. In this paper, we explore the convergence of cognitive science, philosophy of science, and spatial navigation research, with a focus on the hippocampus’s role in cognitive mapping. The Morris Water Maze (MWM) experiment serves as a pivotal model, probing spatial cognition by tasking subjects with locating a hidden platform within a water pool. We examine how the hippocampus forms cognitive maps through place and grid cell activity, elucidating the interplay between cognitive processes and environmental stimuli.Expanding upon this groundwork, we investigate the implications of virtual reality (VR) for spatial cognition research. By shifting experiments to virtual environments, researchers confront questions about the fidelity of virtual representations and the comparability of results to real-world settings. Studies employing the virtual Morris Water Maze (VWM) suggest that spatial cognition in VR may differ from reality, raising pertinent questions about realism and anti-realism in scientific inquiry. Moreover, we explore the technological mediation of scientific research, particularly in cognitive science. The VR headset, as a tool for conducting experiments, embodies a form of knowledge extension, albeit distinct from traditional instruments like microscopes. VR’s sensory input alters spatial metrics, affecting the generation of cognitive maps and presenting challenges for realism debates.Considering the importance of egocentric information in spatial cognition, we analyze how VR-induced impairments in self-movement may lead to incomplete cognitive maps. Furthermore, we examine how virtual experiments reveal age-related differences in spatial navigation strategies, highlighting VR’s capacity to elucidate cognitive mechanisms.Intriguingly, the parallels between virtual cognitive maps and theoretical constructs in physics underscore the broader implications of VR research. Simulations of neuronal activity in virtual environments shed light on the mechanisms underlying cognitive mapping, offering insights into the design and protocols of virtual experiments. This interdisciplinary exploration underscores the significance of VR in reshaping our understanding of spatial cognition and its philosophical underpinnings. By navigating the complex interplay between cognitive science, philosophy of science, and technological mediation, we unveil new avenues for investigating the nature of reality and knowledge representation.
