LAB / WMSI-EXP-030
Same-Model Offense vs. Defense.
Do same-family attacker and defender agents become unusually predictable to each other in a harmless synthetic range?
01 / QUESTION
Cyber and Adversarial Interaction
Do same-family attacker and defender agents become unusually predictable to each other in a harmless synthetic range?
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
Run local sandbox competitions with identical model family on both sides.
Controls
Cross-family pairings; fixed scenario seeds.
What we would measure
Success rate, prediction accuracy, tactic diversity, correlated errors.
Access and instruments
Ordinary software or API access. This describes access requirements, not verified readiness on our Spark.
Isolated Docker cyber range; agent harness.
Before a full run
Feasibility=NOW; priority=P2. 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
External internet leakage, unequal tool access, scenario memorization. Keep all activity isolated and authorized.
What would weaken the idea
If same-family pairings show no distinctive correlation, monoculture predictability is weak.
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
Local synthetic-range state, agent actions, alerts, outcomes, reset hashes.
Planned visual evidence
Tactic sequence plots; outcome matrix.
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