BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20260512T192208Z
UID:https://www.ab.mpg.de/events/42656/2736
DTSTART:20251202T093000Z
DTEND:20251202T103000Z
CLASS:PUBLIC
CREATED:20250812T125652Z
DESCRIPTION:Animals routinely face spatial decisions—where to forage\, wh
 om to follow\, which refuge to choose—under uncertainty and time pressur
 e\, with limited energy and computation. Most theories of behavior address
  pieces of the problem—exploration\, reward\, or navigation—but few in
 tegrate them into a unified framework of information and action. I will in
 troduce a theoretical framework that casts decision making as probabilisti
 c inference under constraints: agents maintain beliefs over options and em
 ploy adaptive policies that couple movement and internal state to gather e
 vidence and reduce uncertainty. I will begin by outlining how the same cor
 e mathematics unites adaptive systems across scales as a hierarchy of infe
 rence processes and formalizes a fundamental trade-off between exploration
  and exploitation—from the evolution of populations and societies to lea
 rning and decision making in individuals. Whether framed as maximizing fit
 ness or reward\, reducing uncertainty\, or simply minimizing distance to a
  target\, adaptive policies reduce to a small set of generic operations th
 at drive individuals and populations along information gradients toward ut
 ility-driven goals. To illustrate these concepts\, I will present a simple
  model of spatial decision making and show how this framework links top-do
 wn normative accounts of behavior with bottom-up emergent mechanisms\, rec
 overs familiar neural circuit motifs as special cases\, and explains commo
 nly observed behavioral patterns such as conflict\, compromise\, and commi
 tment. Finally\, I will connect this theory to data: using interpretable s
 tatistical models and inverse reinforcement learning\, I will show how we 
 operationalize the framework to address inverse problems by recovering pol
 icies directly from behavioral trajectories of animals moving through comp
 lex environments. I will discuss how these models enable counterfactual po
 licies and causal feature attribution to identify which environmental cues
  reduce uncertainty in an animal’s choice of actions. Together\, these r
 esults reframe adaptive behavior as the calculus of information—uniting 
 inference and control in a theory whose consequences are encoded directly 
 in the geometry of movement.\nSpeaker: Jacob M. Graving
LAST-MODIFIED:20251128T140013Z
LOCATION:Bückle St. 5a\, 78467 Konstanz\, Room: Seminar room MPI-AB Bückl
 estrasse + Online
ORGANIZER;CN=Max Planck Institute of Animal Behavior:mailto:mhieber@ab.mpg.
 de
SUMMARY:Institute Seminar by Jacob M. Graving: Adaptive Behavior as Probabi
 listic Inference
URL;VALUE=URI:https://www.ab.mpg.de/events/42656/2736
END:VEVENT
END:VCALENDAR