In artificial intelligence (AI) research, neuralese is a method where large language models (LLMs) perform intermediate reasoning steps in their high-dimensional latent space (vector embeddings) rather than outputting human-readable text tokens. While standard chain-of-thought reasoning forces a model to generate a sequence of words, neuralese allows the model to pass raw, continuous vectors between computational layers, creating a high-bandwidth, non-linguistic reasoning channel.Hao, Shibo, et al. "Training Large Language Models to Reason in a Continuous Latent Space." arXiv preprint arXiv:2412.06769, 9 December 2024.
See also
Artificial intelligence safety Interpretability (machine learning)
References