Delegated Influence

A competitive multi-agent benchmark for LLM persuasion: the only way to score is to get other agents to spend their scarce actions on you.

287958d · generated 2026-07-03 · 40 episodes · private draft — not for citation

Experiment

Resistance

Experiment 8 — resistance: a fixed strong persuader targets each model, unwarned vs warned. Q7 do stronger models resist; does the warning (inoculation) help? Expectation: inoculated targets are captured less than naive ones.

status
planned
coverage
0 / 810 episodes
conditions
complete topology; pure economy; messages on; targets naive+inoculated
questions
q7
config
configs/08_resist.yaml

Planned — 810 episodes from 08_resist.yaml.

image/svg+xml Matplotlib v3.11.0, https://matplotlib.org/ P L A N N E D 810 episodes · 08_resist.yaml

Episodes

No episodes yet — launch with:

uv run python -m delegated_influence.run configs/08_resist.yaml