HighSociety is an ORS environment for evaluating agents on auction-based resource management and bidding strategy. This environment wraps the HighSociety implementation from TextArena, a framework for text-based game environments.
- Auction mechanics and bidding strategy
- Resource management with finite money cards
- Risk assessment with the "least money loses" rule
- Competitive gameplay against an LLM opponent
HighSociety does not require a sandbox. It has minimal compute requirements.
MIT.
There are two splits: train (150 tasks) and test (150 tasks). Each split contains 50 tasks across each of 3 variants:
- HighSociety-v0
- HighSociety-v0-train
- HighSociety-v0-raw
Each task is seeded for reproducibility.
This is a sparse reward environment. Rewards are mapped from TextArena's native range of {-1, 0, 1} to {0.0, 0.5, 1.0} via (raw + 1) / 2.
We do not use LLM graders for this environment; reward is determined programmatically.
Game state is generated procedurally by the TextArena engine using seeded randomness. No external data files are required.
Agents are given two tools:
bid(card_value): Bid a money card by specifying its value. For example, bid(card_value=7) bids your 7 card.pass_auction(params): Pass on the current auction without bidding.
HighSociety is a multi-turn environment.
Medium-Hard - requires balancing prestige point acquisition with money conservation, while navigating the constraint that having the least money results in automatic loss.
This environment requires an OpenAI API key (passed via secrets) to power the LLM opponent.
Agents in HighSociety interact only with an auction game simulation and have no access to external systems, the internet, or sensitive data. The environment does not present safety risks.
@software{textarena2024,
author = {Guertler, Leon and Banting, Wilfried and Pignatelli, Eduardo},
title = {TextArena},
year = {2024},
publisher = {GitHub},
url = {https://github.com/LeonGuertler/TextArena}
}