#deep learning

Mistral AI released Robostral Navigate, an 8B parameter model that enables robots to navigate complex environments using only a single RGB camera and language prompts, achieving 76.6% success on unseen R2R-CE benchmarks. This is significant because it demonstrates map-less navigation, solving the 'kidnapped robot' problem, and outperforms multi-sensor approaches with a simpler setup, potentially enabling wider adoption in hobbyist and commercial robotics. The model is not openly available; it uses only a single camera without pre-captured maps. It achieves state-of-the-art results on the R2R-CE benchmark.

Anthropic researchers have identified a 'global workspace' in large language models (LLMs), where information is integrated across layers, analogous to the Global Neuronal Workspace theory of consciousness. This finding provides a new framework for understanding how LLMs combine information across layers, potentially leading to improved model interpretability and alignment. The work defines a 'J-space' based on the expected change to final logits from perturbations at each layer, revealing a shared subspace across contexts.