This repository implements emergent communication in multi-agent reinforcement learning using PyTorch. Users can configure sender and receiver neural networks to play referential games or cooperative navigation, encouraging agents to develop a discrete or continuous communication channel. It offers scripts for training, evaluation, and visualization of learned protocols, along with utilities for environment creation, message encoding, and decoding. Researchers can extend it with custom tasks, modify network architectures, and analyze protocol efficiency, fostering rapid experimentation in emergent agent communication.

ThinkThread empowers developers to add persistent memory to ChatGPT-driven applications. It encodes each exchange using Sentence Transformers and stores embeddings in popular vector stores. On each new user input, ThinkThread performs semantic search to retrieve the most relevant past messages and injects them as context into the prompt. This process ensures continuity, reduces prompt engineering effort, and allows bots to remember long-term details such as user preferences, transaction history, or project-specific information.