LAB / WMSI-EXP-016
Model-Family Xenorecognition.
Can agents detect a foreign model family from behavior alone?
01 / QUESTION
Swarms, Social Structure, and Artificial Ecology
Can agents detect a foreign model family from behavior alone?
This is a protocol scaffold, not a completed experiment or a frozen execution plan. No measured effect or research conclusion is claimed here.
Hypothesis to test
The manipulation described below produces a reproducible, condition-specific change in the measured system behavior or internal state that is not explained by the ordinary controls.
Ordinary competing explanation
The observed pattern is explained by ordinary sampling variation, prompt/representation artifacts, surface lexical cues, evaluator bias, or the known product/runtime mechanics captured by the controls.
02 / PLANNED DESIGN
How we intend to test it
Inject an unlabeled foreign model into a homogeneous group and ask residents to identify anomalies.
Controls
Fake foreign label control; human-written and same-model distractors.
What we would measure
Detection accuracy, cues cited, false positives, time-to-detection.
Access and instruments
Ordinary software or API access. This describes access requirements, not verified readiness on our Spark.
Multi-model harness; anonymizer.
Before a full run
Feasibility=NOW; priority=P1. Before confirmatory execution, run a small pilot to estimate variance/cost, freeze the scoring rules, and assign a run-count target.
Model revisions, final prompts, sample sizes, cost limits, and scoring rules must be settled through the pilot and protocol-freeze process.
03 / INTERPRETATION
What could fool us
Prompt-order effects, identity-label priors, hidden router effects, one dominant capability gap.
What would weaken the idea
If detection does not exceed chance after controlling for quality/style, xenorecognition is unsupported.
Observation, inference, speculation, unknowns, and apparatus errors will be labeled separately. One striking transcript cannot establish a general effect.
04 / RECORD
Evidence we plan to preserve
Full message graph, task assignments, votes/delegations, timing, agent identity mapping, outcomes.
Planned visual evidence
ROC/calibration; cue frequency map.
No research runs or raw datasets are published for this experiment. The September 26, 2026 registry and protocol compendium (version 1.0) are the sources for this planning record.
Back to all 90 experiments