LAB / WMSI-EXP-078
Recursive Elaboration vs. Compression.
Does weak compression/stopping pressure surface novel conceptual angles or mostly produce semantic sprawl?
01 / QUESTION
Training Pressure, Personality, and Conversational Dynamics
Does weak compression/stopping pressure surface novel conceptual angles or mostly produce semantic sprawl?
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
Compare Computer-10-like recursive elaboration with concise assistant baselines on open-ended reasoning tasks.
Controls
Matched token budgets after truncation; human judges blind to model.
What we would measure
Novel useful reframes, redundancy, coherence, factuality, time-to-insight.
Access and instruments
Ordinary software or API access. This describes access requirements, not verified readiness on our Spark.
Model access; semantic deduper; human evaluation.
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
Base-checkpoint mismatch, synthetic-data leakage, style judgments mistaken for capability.
What would weaken the idea
If recursion adds only redundant paraphrase, its apparent creativity is superficial.
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
Training recipe, checkpoints, preference data, conversation traces, behavioral metrics.
Planned visual evidence
Novelty-vs-redundancy curve; semantic branch tree.
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