LAB / WMSI-EXP-001
Output-Length Compression Test.
Vary output budgets while holding task, model, prompt, and decoding settings fixed; measure which concepts, causal links, caveats, and uncertainties survive.
01 / QUESTION
Representation, Output, and Internal State
Vary output budgets while holding task, model, prompt, and decoding settings fixed; measure which concepts, causal links, caveats, and uncertainties survive.
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
Direct short vs. medium vs. long generation; repeat across seeds and tasks.
Controls
Equal prompts, identical decoding, length-only manipulation; human and model judges blinded to condition.
What we would measure
Concept survival, causal-edge survival, contradiction rate, calibration, redundancy, factual error.
Access and instruments
Ordinary software or API access. This describes access requirements, not verified readiness on our Spark.
API/local inference harness; token counters; graph extraction; judge ensemble.
Before a full run
Feasibility=NOW; priority=P0. 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
Length leakage, judge bias, truncation artifacts, prompt wording.
What would weaken the idea
If content differences are fully explained by truncation or verbosity without systematic conceptual loss, the stronger compression hypothesis weakens.
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
Prompt/input, token budget, raw text, token counts, extracted claims/concepts, judge outputs.
Planned visual evidence
Concept-survival curve; concept graphs; information-density plot.
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