SpecRef hybrid AR-diffusion decoding is tested on six benchmarks with three protocols, showing code benchmarks conflate structural and logical correctness, refinement can degrade correct tokens, and log-likelihood versus generative scoring produce inconsistent model rankings.
and Ben-Nun, Tal and Cardei, Michael and Kailkhura, Bhavya and Fioretto, Ferdinando
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
PoE-Bridge uses a product-of-experts bridge between diffusion and autoregressive distributions, with DLM drafting plus rejection and importance sampling, to deliver 5x speedup over standard DLM decoding while recovering at least 95% of AR performance on math and coding tasks.
SimSD adds a masking strategy to enable speculative decoding in diffusion LLMs, delivering up to 7.46x throughput gains on SDAR models while preserving generation quality.
citing papers explorer
-
Diffusion Language Model Parallel Decoding via Product-of-Experts Bridge
PoE-Bridge uses a product-of-experts bridge between diffusion and autoregressive distributions, with DLM drafting plus rejection and importance sampling, to deliver 5x speedup over standard DLM decoding while recovering at least 95% of AR performance on math and coding tasks.
-
SimSD: Simple Speculative Decoding in Diffusion Language Models
SimSD adds a masking strategy to enable speculative decoding in diffusion LLMs, delivering up to 7.46x throughput gains on SDAR models while preserving generation quality.