Simons Collaboration on Ecological Neuroscience
About SCENE
A hypothesis-driven, multi-species consortium assembling expertise from systems neuroscience, computational neuroscience, cognitive science, machine learning, and developmental psychology.
Mission & Vision
SCENE seeks to uncover general principles of brain function through an ecological perspective: studying perception, cognition, and action in the context of the affordances available to different agents.
Our Definition of "Ecological"
We argue that two complementary approaches are shaping the next decade of systems neuroscience: a naturalistic approach and an ecological one.
- Naturalistic neuroscience emphasizes complex inputs and outputs — for example, fluttering leaves rather than random dot kinematograms, or complex movements rather than saccading to one of two targets.
- Ecological neuroscience emphasizes the structured interactions between the organism and its environment. Artificial inputs and even highly trained behaviours can be ecological, provided they embody the right organism-environment coupling and a computationally-appropriate challenge.
SCENE focuses on the latter. In SCENE, we consider a task to be ecological if it includes three elements:
- A closed perception-action loop between the organism and its environment.
- Sequences of actions with a delayed reward structure, creating a credit-assignment problem in linking actions to outcomes.
- Partial observability, requiring brain states that synthesize sensory inputs over time to infer hidden world variables and guide actions.
This last element — partial observability — marks a key departure from traditional Ecological Psychology and its notion of direct perception. SCENE draws inspiration from Gibson's emphasis on perception-action loops and environmental affordances, but extends it toward mechanistic neurobiological and computational accounts that span species and neural systems.
Our Approach
Ecological Neuroscience
Drawing on Gibson's emphasis on perception-action loops and environmental affordances, SCENE extends ecological psychology toward mechanistic neurobiological and computational accounts. We focus on tasks with closed-loop interactions, sequential decisions, and partial observability — where brains must infer hidden world variables to guide action.
Cross-Species Experiments
SCENE studies mice, bats, non-human primates, and humans — each sensing the world differently (eye-movements, whisking, echolocation) and with distinct bodily effectors. Species are challenged with problems of parallel computational structure — evidence accumulation, affordance identification, exploration vs. exploitation — adapted to each species' abilities.
Theory + Data + AI
Grounded in POMDPs (Partially Observable Markov Decision Processes), SCENE's theoretical framework formalises the closed loop between observations, beliefs, actions, and the environment. A dedicated data science team builds open pipelines for pose estimation, dimensionality reduction, and latent dynamics inference across all species.
Team Science
SCENE has structured its consortium after a "hub-and-spokes" model, with the Theory & Data Analysis team as the hub, and different experimental species (mouse, bat, non-human primate, and human) as different spokes.
This organization allows for multi-tiered and dynamic communication across the collaboration: early and frequent communication within a spoke (~6–15 individuals) and more targeted engagement across spokes (~50–75 individuals) — always couched within the broader theoretical framework.
This model also allows for increased scientific ownership of projects within each spoke, and for each species to operate within its appropriate niche. Both within and beyond our field, there is a growing appreciation for how team-science can undertake projects a single lab cannot accomplish alone, and how it fosters multi-disciplinarity and enhances scientific rigor through the early and frequent sharing of data, code, and protocols.
Funders & Institutions
SCENE (opens in new tab) is part of the Simons Collaborations in Neuroscience (opens in new tab). The work is supported by a Simons Foundation International grant administered by the Simons Foundation.