← Back to Catalog
The T-Maze Problem

A rat faces a T-shaped maze. At the junction, it must choose: go left or go right? One side has cheese 🧀, the other an electric shock ⚡.

But first it can visit a cue location that reveals which side is safe. This simple setup reveals the Coherence-Rupture-Regeneration dynamics of decision-making.

C → δ → R

The rat accumulates coherence (builds certainty through observation), experiences rupture (the decision moment), and regenerates (selects action based on memory-weighted integration).

CRR treats each decision as a scale-invariant process: the agent metabolizes its accumulated past into future action at the rupture point δ(now).

The cue location is where coherence accumulates. Before visiting, the rat has low coherence about context — uncertainty about which side is safe.

After visiting the cue, coherence is high: the rat now has integrated, consistent knowledge about its environment.

Coherence Integral
C(x,t) = ∫ L(x,τ) dτ

L represents the local coherence field — how consistent the agent's state is with its environment at each moment.

  • Before cue: C ≈ 0.5 (uncertain, incoherent)
  • After cue: C ≈ 0.98 (certain, coherent)
  • Each observation integrates into the coherence field

CRR uses a generative process with explicit temporal dynamics:

Coherence Field

How certain the agent is about context. Observation builds coherence; uncertainty depletes it.

Ω (Omega) — Boundary Parameter

Controls how memory is accessed. Low Ω = only recent high-coherence moments. High Ω = broad historical access.

Memory Kernel exp(C/Ω)

Weights past states by coherence. High coherence moments dominate regeneration when Ω is small.

Reward History

Past experiences of reward/punishment, weighted by coherence at time of experience.

At each rupture point δ(now), the rat evaluates possible actions via Regeneration Potential (R):

Regeneration
R = ∫ φ(x,τ) · exp(C/Ω) · Θ(...) dτ

This decomposes into two natural components:

Coherence-Building Value

How much visiting a location will increase coherence. The cue has high coherence-building value when uncertainty is high.

Memory-Weighted Preference

Expected outcome weighted by exp(C/Ω). Higher coherence about reward location → stronger pull toward that action.

CRR naturally balances "information seeking" and "reward seeking" through the exp(C/Ω) weighting — both serve coherent regeneration.

Ω (omega) controls boundary flexibility — how the agent accesses its memory during regeneration:

Policy Selection
P(π) ∝ exp(R(π) / Ω)

Low Ω: Only highest-coherence memories accessible. Peaked, decisive action selection. The agent acts on its most certain knowledge.

High Ω: Broad memory access. More exploratory, considers diverse possibilities. The agent remains open to alternatives.

  • Low Ω → Habitual, pattern-reconstituting
  • High Ω → Adaptive, exploratory
  • Contemplatives train Ω control through practice

Watch the rat's behaviour emerge from CRR dynamics. With low initial coherence, the cue has high coherence-building value → the rat visits it first.

After building coherence at the cue, the memory-weighted preference for the correct side dominates → the rat goes to the cheese.

"Exploration vs exploitation" dissolves in CRR: both are aspects of coherent regeneration. Information-seeking builds coherence; reward-seeking realizes it.

Adjust the sliders to see how coherence weight, preference weight, and Ω affect the rat's choices.

CRR Framework — Sabine (2024-2025)
Coherence-Rupture-Regeneration dynamics
CRR T-Maze Simulation
4.0
1.5
6
Coherence Field (Context)
LEFT
50%
RIGHT
50%
Regeneration Potential (R)
CUE
LEFT
RIGHT
REWARDS:0
SHOCKS:0
TRIALS:1
STEP:1

CRR Dynamics in Phase Space

Coherence and Regeneration Attractors
Coherence Attractor
Regeneration Attractor
Agent State
Flow Field

CRR Parameters

Current certainty about context (0 = incoherent, 1 = fully coherent)
LOW HIGH
20%
Importance of building coherence (information value)
LOW HIGH
1.0
Importance of memory-weighted reward (regeneration pull)
LOW HIGH
1.0
Boundary flexibility: Low = peaked/decisive, High = broad/exploratory
RIGID FLEXIBLE
Ω = 2.0

Live CRR Decomposition

Coherence pull:
0.00
Regen. pull:
0.00
Combined R:
0.00
🌀

Low coherence → agent seeks information first (coherence attractor dominates).

"At rupture, the agent metabolizes accumulated coherence into future action. The exp(C/Ω) weighting ensures that high-coherence moments dominate regeneration — we act from what we know most clearly."

— CRR Framework: Coherence-Rupture-Regeneration

CRR Glossary

Coherence-Rupture-Regeneration concepts
Coherence C(x,t)
Accumulated consistency between agent and environment. C = ∫L(x,τ)dτ. High coherence = clear, integrated knowledge.
Rupture δ(now)
Scale-invariant choice moment. The present instant where past becomes future. Every decision is a tiny "now" where you commit.
Regeneration R
Memory-weighted integration toward action: R = ∫φ·exp(C/Ω)·Θ dτ. You regenerate from your clearest memories.
Ω (Omega)
Boundary flexibility parameter. Controls memory access breadth in exp(C/Ω). Low Ω = peaked focus. High Ω = broad access.
exp(C/Ω)
Memory amplification kernel. Weights past states by coherence. Clear memories dominate; fuzzy ones fade.
Coherence Building
Actions that increase C — reduce uncertainty, integrate information. Checking the cue builds coherence about context.
Local Coherence L(x,τ)
Instantaneous consistency at location x and time τ. How "right" does this moment feel?
Reconstruction φ(x,τ)
Resources available for regeneration at each past moment. What can you build from each memory?
Policy Selection
P(π) ∝ exp(R(π)/Ω). Actions selected by regeneration potential. Do what regenerates you best.
Z₂ Symmetry
Binary switching systems: Ω = 1/π, CV ≈ 0.159. On/off, left/right, yes/no.
SO(2) Symmetry
Continuous rotation systems: Ω = 1/2π, CV ≈ 0.08. Cycles, rhythms, oscillations.
Metabolization
At δ(now), past coherence is transformed into future action. You digest experience into behavior.
CRR Framework | Alexander Sabine | cohere.org.uk | temporalgrammar.ai