Skip to content

Consciousness-Indicator Architecture (CIA)

Cave wall warning: CIA only measure theory signals. CIA not prove real inner feeling, sentience, or magic mind light.

What this package be

CIA be Python tool for research tribe. Tool check architecture signs that maybe match consciousness ideas. Tool run modular think-loop and make scorecard on 0-22 scale.

Good for AI safety tribe, cognitive science tribe, and mind-philosophy tribe when tribe want clear testable proxy numbers.

Big things tool do

  • 11 indicator groups from known consciousness theories
  • Full cognitive cycle simulation
  • Scorecard output with risk tiers
  • Causal intervention tests (ablation style)
  • Governance checks and report verification tools
  • Optional LLM adapters (cloud and local)
  • Optional advanced_cia research parts (interaction model, on-policy distillation, LoRA, manifold constraints, deterministic inference)
  • Optional EEG/BCI neuroadaptive part
  • Optional plant biohybrid part

Install

From source

pip install -e .

Use ConsciousnessAI 0.0.1 for non-advanced features.

With extras

pip install -e ".[dev]"
pip install -e ".[dev,neuro,plant,llm,local-llm]"

Python need

  • Python 3.11+

Command quickstart

cia run "A red object moved behind a screen and reappeared."
cia run "The system noticed its own processing limits." --verbose
cia score --input examples/sample_inputs.json
cia intervene --type disable_workspace

Python quickstart

from cia.simulation import CombinedConsciousnessIndicatorSystem

system = CombinedConsciousnessIndicatorSystem(recurrent_cycles=3)
report = system.run_cycle("The agent reflected on its own uncertainty.")

print(report.indicator_scores.total_score, report.indicator_scores.max_possible)
print(report.welfare_state.risk_level)

Optional integrations

LLM

  • Providers: OpenAI, Claude, Gemini, Hugging Face, llama.cpp, vLLM
  • Adapter load from config

Neuroadaptive EEG/BCI

  • Offline ingest and preprocessing
  • Feature extraction and proxy-state conditioning

Boundary: EEG features be statistics only. No mind-reading.

Plant biohybrid

  • Non-invasive plant electrophysiology ingest
  • Signal-feature conditioning for CIA runtime control

Boundary: plant signals be physiology proxy only, not proof of plant consciousness.

Docs and reports

  • Full project docs: docs/ in repo
  • Validation/demo outputs: reports/ in repo
  • Citation data: CITATION.cff

For full usage, benchmarks, governance policy, and scientific caveats, see MkDocs site from this repo.

License

MIT License