Persistent Latent Misalignment, a new dimension of misalignment?
A new paper was released at ICML that I'm worried will open an entire new dimension of alignment problems:Latent Collaboration in Multi-Agent Systems (LatentMAS)TLDR: they show that multiagent systems can communicate faster and more efficiently by directly sharing latent states instead of text. This is a large performance improvement that comes at the cost of interpretability.Previously, the handover between agents was a bottleneck where latent thoughts had to be put into text. Steganography was...
Read full article →