[KongchangAI]
ConceptOut-of-Distribution Damping

OOD Damping

A technique that penalizes samples deviating too far from the natural distribution during generation, used to control the plausibility of generated sequences while preserving room for exploration.

Timeline (last 90 days)

Nov 2

LOCOGEN contains two key components: invariant motif mask learning (identifying motif positions that are critical to function and protecting them from excessive guidance) and OOD damping (penalizing generations that deviate too far from the natural distribution at background positions, but acting only on background regions).

Unverified50%

All Facts (1)

Source Articles