cognize 0.1.3

Symbolic cognition engine for epistemic drift, rupture detection, and realignment.

Cognize

Programmable cognition for Python systems

---

Overview

Cognize is a lightweight cognition engine for Python systems.
It tracks belief (V) vs. reality (R), manages misalignment memory (E), and detects symbolic rupture (Θ).
Now supports runtime injection of programmable logic for collapse, realignment, and adaptive thresholds.

Built for agents, simulations, filters, and symbolic drift-aware systems.

---

Features

• Cognitive projection engine (EpistemicState)
• Drift tracking with misalignment memory
• Programmable inject_policy(...) support
• Prebuilt logic in cognize.policies (collapse, realign, threshold)
• Vector-compatible input support
• Trace export (.json, .csv) for audit or training
• Lightweight, domain-agnostic, DSL-ready

---

Installation

pip install cognize

---

Core Concepts

Symbol	Meaning
V	Belief / Projection
R	Reality Signal
∆	Distortion
Θ	Rupture Threshold
E	Misalignment Memory

---

Quick Usage

from cognize import EpistemicState

# Initialize agent
agent = EpistemicState(V0=0.0, threshold=0.35)

# Feed signals
for R in [0.1, 0.3, 0.6, 0.8]:
    agent.receive(R)
    print(agent.summary())

---

Programmable Injection (v0.1.2+)

from cognize.policies import (
    collapse_soft_decay_fn,
    realign_tanh_fn,
    threshold_adaptive_fn
)

agent.inject_policy(
    collapse=collapse_soft_decay_fn,
    realign=realign_tanh_fn,
    threshold=threshold_adaptive_fn
)

agent.receive(0.5)
print(agent.summary())

---

Exporting Trace

agent.export_json("trace.json")
agent.export_csv("trace.csv")

---

Example Output

{
  "t": 2,
  "V": 0.41,
  "R": 0.6,
  "delta": 0.19,
  "Θ": 0.35,
  "ruptured": false,
  "symbol": "⊙",
  "source": "default"
}

---

Full Cognize User Guide

---

License

Licensed under the Apache License 2.0.

---

© 2025 Pulikanti Sashi Bharadwaj
All rights reserved.