Neural Thickets gains traced to output diversity rather than weight experts
Perturb-Select-Vote on frozen models matches or beats full-weight perturbations on most reasoning tasks and helps vision under domain shift.
2 stories, newest first · one card per story
Perturb-Select-Vote on frozen models matches or beats full-weight perturbations on most reasoning tasks and helps vision under domain shift.
Researchers audit 31 NLP techniques for clinician-annotated suicide risk prediction, finding only 5 of 31 comparisons yield reliable gains.
Today 1.6K readers · 2.4K views · 775 reads · all time 189K views