← back to timeline

EXP-DOSE80

informative

Integrative Dose-Response — 80% (4/5 Synthesizers)

2026-03-10 L2 level paper3 4 runs $0.02
new
premortem
mock
real
metrics
analyze
review
krit
done

In plain language

Four out of five agents are synthesizers. The sole remaining analyst is like one source among four consumers. Will it dominate thanks to scarcity, or will four synthesizers overwhelm it?

What we found

No predictions scored

Predictions we made before running

0/4 confirmed

Technical details

Research hypothesis

DOSE-RESPONSE SERIES: 0% → 20% → 40% → 60% → [80%] → 100%

Prior data points (TR range):

  • 0% (EXP-PENT): 0.528
  • 20% (EXP-ENGINEER): 0.240
  • 40% (EXP-DOSE40-SINK): 0.404
  • 60% (EXP-DOSE60): 0.352

Pattern: DAMPED OSCILLATION with decreasing amplitude: 0.528 →(-0.288)→ 0.240 →(+0.164)→ 0.404 →(-0.052)→ 0.352

At 80% (4 synthesizers, 1 non-integrative), Alpha is the SOLE remaining analytical agent. Four synthesizers compete for Alpha's upstream vocabulary.

Two competing predictions: (a) Damped oscillation continues: small positive swing → TR ≈ 0.37-0.38 The oscillation has been converging, amplitude ~0.03 for the next half-cycle. (b) Lone analytical agent gets AMPLIFIED: one unique source among 4 identical synthesizers → Alpha becomes extreme upstream → TR INCREASES to >0.40. Scarcity of non-integrative vocabulary makes it more valuable.

The varied-rotation design (4 rotations) tests whether position confounds drive individual placement, per EXP-ENGINEER-REPL learning.

Experimental setup

Type: single_condition

Condition Parameters
PERSONA_LIVE persona: PERSONA, interaction: LIVE

Factors: persona (PERSONA) × interaction (LIVE)

Parameters

n_agents
5
n_rounds
50
n_runs_per_condition
4
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).