NeoDiff uses a Poisson forward process with per-token noise timing and a learned time predictor to improve non-autoregressive text diffusion generation.
Evaluate this paraphrase generation (0-100 score): [Original] {source} [Reference] {reference} [Paraphrase] {paraphrase} Score these aspects STRICTLY IN THIS ORDER:
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Unifying Continuous and Discrete Text Diffusion with Non-simultaneous Diffusion Processes
NeoDiff uses a Poisson forward process with per-token noise timing and a learned time predictor to improve non-autoregressive text diffusion generation.