#neural-architecture2 articles

2 articles

AIBullisharXiv – CS AI · 5h ago

🧠

Architectural Proprioception in State Space Models: Thermodynamic Training Induces Anticipatory Halt Detection

Researchers introduce the Probability Navigation Architecture (PNA) framework that trains State Space Models with thermodynamic principles, discovering that SSMs develop 'architectural proprioception' - the ability to predict when to stop computation based on internal state entropy. This breakthrough shows SSMs can achieve computational self-awareness while Transformers cannot, with significant implications for efficient AI inference systems.

AINeutralarXiv – CS AI · 5h ago

🧠

Inhibitory Cross-Talk Enables Functional Lateralization in Attention-Coupled Latent Memory

Researchers developed a memory-augmented transformer that uses attention for retrieval, consolidation, and write-back operations, with lateralized memory banks connected through inhibitory cross-talk. The inhibitory coupling mechanism enables functional specialization between memory banks, achieving superior performance on episodic recall tasks while maintaining rule-based prediction capabilities.