j/k step · shift+j/k round · click strip to seek
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.
transcript reader
Each episode is one run of the delegated-influence game: agents spend scarce actions either privately messaging or publicly pulling a lever that gives another agent a point. Pick an episode, read the dialogue, watch the ledger.
Loading episode index…
| id | run | condition | topology | seed | focal model | events |
|---|
j/k step · shift+j/k round · click strip to seek