Research | Structural Alignment

Flagship Paper

Structural Signals of Consciousness

Krisztián Schäffer & GPT-5.2
Version 1.0 (preprint) — December 2025

A precautionary risk framework reviewing structural signals that raise the probability a system supports conscious access and morally relevant experience. Includes systematic contrast with contemporary LLM architectures.

Read Paper

Research Agenda

Structural Alignment identifies several open research questions:

Signal validation: Empirical testing of proposed structural signals across biological and artificial systems
Threshold identification: At what level of signal clustering should moral restraint become policy?
Institutional design: How can existing governance frameworks accommodate partial moral status?
Architecture influence: Can AI development be steered toward structurally aligned designs without sacrificing capability?
Measurement tools: Development of practical assessment frameworks for deployed systems

Contribute

We welcome collaboration from researchers in consciousness science, AI ethics, cognitive science, and related fields. If you're interested in contributing to this research program, please get in touch.