EXP-QUAD2

✓ success

Quartet Parity Isolation — N=4 with Non-Integrative Personas

2026-03-09 L3 level paper3 1 runs

new

premortem

mock

real

metrics

analyze

review

krit

done

In plain language

Our quartet experiment showed flat hierarchy, but was it the even number of agents or the storyteller persona blurring the ranks? We rerun four agents with only non-integrative roles — the same skeptic and analyst who produced hierarchy at five — to find out which variable matters. Answer: the storyteller was flattening the system. Four agents hierarchy fine without one.

What we found

N=4 hierarchy RESTORED (TR 0.509) with non-integrative personas. Odd/even parity REFUTED — integrative persona was the confound.

Predictions we made before running

1/4 confirmed · 3 refuted

▶ Technical details

Research hypothesis

EXP-QUAD (N=4): flat (TR range 0.059). Delta was narrative/synthesis — INTEGRATIVE. EXP-PENT (N=5): hierarchy (TR range 0.528). All personas NON-integrative.

CONFOUND: EXP-QUAD's flatness could be caused by: A) Even-cycle round-robin (topology → symmetry → flat) B) Integrative persona (synthesis role → homogenizer → flat)

This experiment ISOLATES the variable: N=4 with the SAME non-integrative personas as EXP-PENT (minus Epsilon). If flat → parity mechanism confirmed. If hierarchical → the integrative persona was the confound.

This is the CRITICAL disambiguation experiment for the odd/even hypothesis.

Experimental setup

Type: simple

Condition	Parameters
QUAD2_LIVE	persona: PERSONA, interaction: LIVE, n_agents: 4, note: Same as EXP-PENT but N=4 (drop Epsilon)

Factors: n_agents (4) × persona_type (non_integrative)

Parameters

n_agents: 4
n_rounds: 48
n_runs_per_condition: 1
model: gemini-2.5-flash-lite
temperature: 0.9
scheduler: round_robin

Trophic Ratios by Condition

Mean trophic ratio per agent across runs. Error bars = ±1 std dev. Higher TR = more upstream (exporter).