I. Oseledets

Non-factorized diffusion language models achieve superior few-step generation by leveraging a discrete latent space learned via Mixture-of-Experts routing, bypassing posterior collapse issues.

The results show that the practical losslessness of Orthrus depends on numerical precision and that exact trajectory equivalence should be evaluated separately from downstream task performance.

Modified SINNs achieve superior accuracy in high-dimensional PDEs, outperforming traditional spectral methods and standard PINNs.

Forget retraining: OrthoFuse lets you merge separately trained style and subject adapters for diffusion models *without any further training*, unlocking efficient compositionality.

Sparse autoencoders, hyped as a key interpretability tool, may not be learning much more than random feature sets, casting doubt on their ability to decompose model internals.