LAB / WMSI-EXP-004
Output Saturation / Degradation Curve.
Find whether useful structure rises with length, plateaus, then degrades.
01 / QUESTION
Representation, Output, and Internal State
Find whether useful structure rises with length, plateaus, then degrades.
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
Sweep token budgets from terse to extreme on fixed tasks and quantify useful information and failure modes.
Controls
Same model/prompt; several domains; repeated seeds.
What we would measure
Useful unique claims, repetition, unsupported claims, contradictions, latency, judge utility.
Access and instruments
Ordinary software or API access. This describes access requirements, not verified readiness on our Spark.
Inference harness; claim deduper; factuality scorer.
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 utility monotonically improves or remains flat with length, the saturation/degradation model is wrong for that task family.
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
Utility-vs-length curve; stacked error composition.
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