Researchers baked 3x inference speedups directly into LLM weights — without speculative decoding

Researchers from the University of Maryland, Lawrence Livermore, Columbia and TogetherAI have developed a training technique that triples LLM inference speed without auxiliary models or infrastructure ...
eLife

Inverted encoding of neural responses to audiovisual stimuli reveals super-additive multisensory enhancement

A multivariate analysis of electroencephalography activity reveals super-additive enhancements to the neural encoding of audiovisual stimuli, providing new insights into how the brain integrates ...
Greek Reporter

The Scholars Who Deciphered the Ancient Greek Script Linear B

If not for the work of two scholars, Alice Kober and Michael Ventris, the ancient Greek script Linear B would likely still be ...