LAB / WMSI-EXP-085
Model-Routing Transparency Audit.
When a user selects a model/reasoning level, what backend model and effort actually execute each turn?
01 / QUESTION
System and Product Behavior
When a user selects a model/reasoning level, what backend model and effort actually execute each turn?
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
Capture HAR/API metadata in controlled fresh chats across models, effort levels, times, and prompt classes.
Controls
Manual selection vs Auto; fresh sessions; usage-limit state recorded.
What we would measure
Requested/default/resolved slugs, transitions, latency, error/fallback notices.
Access and instruments
Ordinary software or API access. This describes access requirements, not verified readiness on our Spark.
HAR parser; redaction tool; metadata analyzer.
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
Caching/history records mistaken for live execution, hidden fallback notices, client-version mismatch.
What would weaken the idea
If slugs always match selections under controls, silent-routing claims weaken for tested conditions.
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
Client version, account tier, selected mode, request/response metadata, timestamps, screenshots/HAR with secrets redacted.
Planned visual evidence
Routing timeline; transition Sankey; mismatch table.
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