#artificial intelligence
Jun 2026
Attention-Discounted Adaptive Sampler for Masked Diffusion Language Models
ADAS is proposed, a training-free reranking rule that leaves the base sampler's stopping rule unchanged and greedily discounts each token-wise confidence score according to its attention to already selected positions, weighted by their prediction uncertainty.
Y. Şahin, Ahmed R. Saikia, V. Cevher et al.
· arXiv.org · 1 citation