EXP-ABSORPTION-DIAGNOSTIC

… pending

Absorption diagnostic — prompt-level or substrate-level?

2026-04-17 L5 level paper3 15 runs $0.24

new

premortem

mock

real

metrics

analyze

review

krit

done

In plain language

This experiment tested whether a model's tendency to agree with instructions is a general trait or specific to the way the instruction is phrased. Researchers found that a specific type of instruction, designed to make the model disagree, did not change the model's agreement, suggesting this is a general tendency. The results are still being analyzed to understand the exact nature of this behavior.

▶ Technical details

Research hypothesis

Deacon COUPLED showed d_between=-3.39 (convergence, not complementarity). This experiment determines whether that's prompt-level (MINIMAL prompts default to assent) or substrate-level (any prompt converges). H1: CONTRAST instruction breaks absorption → prompt-level H2: CONTRAST fails too → substrate-level → compute-gift motivated H3: PERSONA partial, CONTRAST further → intermediate

Experimental setup

Type: simple

Condition	Parameters
MINIMAL_C	persona: IDENTICAL, coupling: COUPLED, n_runs: 5
PERSONA_C	persona: PERSONA, coupling: COUPLED, n_runs: 5
CONTRAST_C	persona: IDENTICAL, coupling: COUPLED, n_runs: 5

Factors: coupling (COUPLED)

Parameters

n_agents: 6
n_rounds: 80
n_runs_per_condition: 5
model: google/gemini-2.5-flash-lite
temperature: 0.7
scheduler: round_robin

absorption diagnostic coupling prompt-sensitivity endogeneity-ladder-1a-prime