Beta

Explore
Marketplace
Neural Labs
Playground
Wallet
Docs

Beta

Explore Marketplace Neural Labs Wallet Docs

Build

Manage

My assets Manage wallet Organization

Community

Discord X (Twitter)Telegram

Docs Privacy Policy Terms and Conditions

Proud member of

Proud member of

Docs Cookie Policy

Privacy Policy Terms and Conditions

© 2026 AlphaNeural AI. All rights reserved.

cogames-lstm-arena – AI Model by anoop4bhat | AlphaNeural AI

You can deploy this model and start earning money today!

anoop4bhat/
cogames-lstm-arena

like

0

Views

No views yet

CoGames CvC LSTM Policy

Trained on the arena mission of CoGames - the Alignment League Benchmark.

Performance

Average Score: 0.0000 (per cog)
Individual Scores: []
Training Steps: 100,000,000
Policy Architecture: LSTM (341K params, hidden_size=128)
Training Algorithm: PPO (PufferLib)

Hyperparameters

Learning Rate: 0.00092 (annealed)
Gamma: 0.995
GAE Lambda: 0.90
Clip Coefficient: 0.2
Entropy Coefficient: 0.01
BPTT Horizon: 64
Value Function Coefficient: 2.0

Usage

bash

pip install cogames

# Play with this policy
cogames play -m arena -p class=lstm,data=model.pt

# Evaluate
cogames eval -m arena -p class=lstm,data=model.pt

# Submit to leaderboard
cogames create-bundle -p class=lstm,data=model.pt -o submission.zip
cogames upload -p submission.zip -n my-policy
cogames submit my-policy --season beta-teams-tiny-fixed

About CvC (Cogs vs Clips)

A cooperative territory-control game where teams of AI agents work together to:

Gather resources and deposit them at controlled junctions
Acquire specialized roles (Miner, Aligner, Scrambler, Scout)
Capture neutral and enemy junctions
Defend territory from automated Clips expansion