Public evidence and explanation
Publish enough to test the observation. Never publish beyond the evidence.
FASO’s proposed outputs are descriptive and non-prescriptive. Every published finding should identify its evidence, authority, confidence, limitations and publication status.
Current publication register
Every publication carries an explicit evidence status, version and limitation.
The Research and Publications section is the authoritative home for FASO working hypotheses, proposed validation protocols and later findings. Social-media summaries must link back to the complete version.
Publication method · 27 July 2026
The Project FASO Research and Publication Method
The classification, evidence, versioning, correction and publication rules governing FASO research outputs.
Read the complete publicationWorking hypothesis · 27 July 2026
From Instruction Attenuation to Operating-Regime Reversion
A testable hypothesis concerning abbreviation-induced semantic drift and the loss of an accuracy-governing instruction.
Read the complete publicationWorking hypothesis · 27 July 2026
Evaluation-Induced System Drift
A testable hypothesis concerning continual evaluation, persistent feedback and change in either the system being measured or the validity of its measurement.
Read the complete publicationWorking hypothesis · 27 July 2026
Cross-Model Evaluation Propagation and Multiplier Effects
A testable hypothesis concerning the lateral, stack-wide, institutional and generational propagation of evaluation-derived knowledge.
Read the complete publicationValidation protocol · 27 July 2026
Controlled Validation Protocol for Instruction Attenuation
A proposed experiment comparing absent, abbreviated, paraphrased, repeated and mechanically verified governing instructions.
Read the complete publicationThe publication promise
What was observed, how it was verified, what confidence is warranted and what remains unknown.
Public output must be useful enough to support scrutiny and bounded enough to protect people, systems, confidential information and the integrity of the method.
Observation records
Bounded accounts of safety-relevant discovery, change, drift or supported system-boundary departure.
Evidence and replay demonstrations
Controlled artefacts showing provenance, time order, replayability and the formation of results.
Research and technical notes
Methods, limitations, open issues, architecture summaries and validation materials.
Independent findings
External reviews, validation reports, qualifications, negative findings and FASO’s separately identified response.
Institutional reports
Governance, funding, conflicts, corrections, annual activity and publication-boundary decisions.
Publication Boundary
Transparency is disciplined disclosure, not unrestricted release.
Designed for publication
- Bounded system status and dated checkpoints
- Method and governance summaries
- Controlled evidence and replay demonstrations
- Independent findings and limitations
- Corrections and superseded positions
Held behind the boundary
- Protected implementation logic
- Personal, confidential or privileged information
- Unsafe operational detail
- Unresolved or unauthorised evidence
- Conclusions the evidence cannot support
Current public position
Research outputs remain distinct from system evidence and future publication commitments.
Use the complete research register for current publications, or review the evidence register for the built system’s current status.