[AINews] Agents disprove Polysemanticity; greatly improve Interpretability & Alignment

New research challenges the superposition theory of polysemanticity, suggesting neural networks are more interpretable than previously believed through recursive generalization patterns.

Researchers introduce OmniVerifier-C1, a decomposed graded verifier that provides continuous scoring instead of binary judgments for multimodal verification tasks.

Researchers introduce T2-z, a method for test-time optimization of environment latents in robotic world models that improves prediction accuracy and action success rates across simulations and real-world tasks.

Nick Cammarata argues neural networks are already clean and interpretable, challenging previous pessimism about mechanistic interpretability.

Researchers propose MirrorMoE to address the exploration dilemma in modern Mixture-of-Experts architectures between fine-grained segmentation and shared experts.