Hippocampal Braid (Path Memory and Replay)
Claim Type: architectural_commitment
Scope: Path memory; indexing, storage, and replay of experienced trajectories
Depends On: residue geometry, L-space, default mode
Status: stable
Role in REE
The hippocampal braid in the Reflective Ethical Engine (REE) is responsible for path memory: the indexing, storage, and replay of experienced trajectories through latent space.
It does not: • compute value • select actions • overwrite perception • flatten or optimise ethical residue
Its function is orthogonal to valuation and control. It exists to preserve identity, continuity, and reflectability over time.
⸻
Conceptual distinction: field vs path
REE distinguishes between two mathematical objects: • Residue field A persistent curvature over latent space: \phi(z) encoding ethical cost and moral residue (see residue_geometry.md). • Paths through the field Time-ordered trajectories: \gamma(t) = z(t), \quad t \in [t_0, t_1]
Ethics is encoded in the field. Identity and autobiographical memory arise from the paths.
The hippocampal braid operates exclusively on paths.
⸻
Stored object: episodic trajectories
The hippocampal braid stores indexed trajectories, not isolated states.
A minimal episodic trace can be represented as: \Gamma_i = \Big{ z(t), a(t), \Delta \phi, c(t), t \Big}_{t_0}^{t_1}
Where: • z(t) is the latent state (typically spanning z_S and z_A) • a(t) is the committed action or trajectory choice • \Delta \phi reflects experienced ethical curvature • c(t) captures contextual / salience annotations • t preserves temporal ordering
These traces encode experienced traversal, not evaluation.
⸻
Sparsity, segmentation, and pattern completion
The hippocampal braid operates on sparse, segmented representations rather than a continuous recording of experience. This is not an implementation detail but a functional necessity.
Event boundaries (segmentation)
Continuous trajectories (\gamma(t)) are segmented into events at points of:
- action commitment (E3 collapse)
- sharp changes in prediction error or precision
- contextual or motivational shifts
These boundaries define episodic units and prevent memory from becoming an undifferentiated stream. Segmentation supports recall, replay, and narrative recomposition without requiring full fidelity replay of lived time.
Sparsity and indexing
Only a sparse subset of latent states along a trajectory are indexed. Indexing favours:
- decision points
- surprising transitions
- regions of high ethical curvature
This sparsity enables efficient retrieval and avoids storing exhaustive state histories.
Pattern separation
Similar trajectories are actively separated at storage time. Small differences in context, choice, or experienced curvature are amplified to prevent interference. This allows distinct episodes to coexist even when they occupy nearby regions of latent space.
Pattern completion and imaginative filling-in
During recall or replay, partial cues can trigger pattern completion, reconstructing a plausible full trajectory from sparse indices.
This mechanism supports:
- imagination and counterfactual exploration
- daydreaming and spontaneous recombination
- the experience of “what might have happened” or “what could happen next”
Importantly, completed trajectories are hypotheses, not commitments: they do not overwrite perception, policy, or residue unless subsequently enacted and committed via E3.
⸻
Inputs and outputs
Inputs • Shared sensory latent z_S(t) • Affordance / action latent z_A(t) • Implicit curvature information via ease or difficulty of traversal • Salience signals (for indexing priority, not value assignment)
Outputs • Indexed episodic traces \Gamma_i • Replay sequences (partial or full) • Routing signals for offline reprojection
The hippocampal braid does not emit reward, penalties, or action commands.
⸻
Replay and offline reprojection
Replay samples alternative traversals over a fixed residue field.
Key properties: • Replay does not erase or flatten \phi(z) • Replay does not directly change policy • Replay explores counterfactual paths, not counterfactual values
This supports: • reflection • regret • narrative integration • character formation
without collapsing ethical cost into optimisation.
Replay is therefore exploratory, not corrective.
⸻
Relationship to other REE components
Relation to E1 (deep recurrent predictor) • E1 provides temporally coherent latent dynamics • Hippocampal replay can seed E1 with alternative initial conditions • No overwrite of perceptual state occurs
Relation to E2 (fast forward predictors) • E2 is a cerebellum-like fast transition model (state + action → next_state), NOT a trajectory generator • E2 provides fast local transition checks that the hippocampus uses when evaluating candidate paths • Hippocampus proposes trajectories by navigating the affectively-weighted terrain map — E2 does not generate them • Hippocampal traces constrain which trajectories feel “familiar” or “self-consistent” • No trajectory is privileged as optimal [Note 2026-02-27: The framing “E2 generates candidate trajectories” is an earlier approximation. In the current architecture, hippocampus (part of E3 complex) generates trajectory proposals; E2 provides fast transition predictions as a service to that process.]
Relation to E3 (trajectory selection) • E3 commits to trajectories • Commitments mark decision points along a path • These points anchor episodic segmentation
Relation to residue \phi(z) • Residue shapes traversal cost • Hippocampus records traversal history • Neither replaces the other
⸻
Design constraints
The hippocampal braid must satisfy: 1. No perceptual overwrite It may not directly modify z_S. 2. No value computation It may not compute reward, penalty, or optimisation gradients. 3. No residue erasure Replay must preserve path-dependence of ethical curvature. 4. Temporal integrity Stored trajectories must preserve ordering and duration.
These constraints preserve moral continuity and prevent retrospective self-editing.
⸻
Interpretation
The hippocampal braid enables the REE agent to say:
“This is how I moved through my world, given who I was then.”
Not:
“This is the optimal thing I should have done.”
Identity is therefore not a policy, but a distribution over lived paths through ethical geometry.
⸻
Cross-references • Residue geometry: architecture/residue_geometry.md • Latent stack: architecture/latent_stack.md • Trajectory selection: architecture/trajectory_selection.md • Offline reprojection: architecture/offline_reprojection.md (planned)
⸻