The paper introduces Uni-E, a unified energy for DLMs that accounts for model capacity, dependency and invariance, can be computed exactly, and corrects distribution shifts from dependency and invariance.
Self speculative decoding for diffusion large language models
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
years
2026 9verdicts
UNVERDICTED 9roles
method 1polarities
extend 1representative citing papers
FeF-DLLM achieves factorization-error-free generation in discrete diffusion language models via prefix-conditioned posterior factorization and speculative decoding, delivering 5.04 pp higher accuracy and 3.86x faster inference on GSM8K, MATH, HumanEval, and MBPP.
A3 reframes dynamic action chunk commitment in VLA models as self-speculative prefix verification, accepting the longest continuous sequence of actions that satisfies consensus-ordered conditional invariance and prefix-closed sequential consistency.
MARS fine-tunes autoregressive models to predict multiple tokens per step via continued training on instruction data, achieving 1.5-1.7x throughput while matching baseline accuracy and supporting real-time speed adjustment.
TEAM accelerates MoE dLLMs up to 2.2x by exploiting temporal-spatial consistency in expert routing to accept more tokens with fewer activations.
A new residual-sampling scheme for diffusion models permits block verification and yields up to 6.3% speedup via a heuristic self-speculative drafter that needs no training.
AGDO improves dLLM reasoning performance by determining denoising order and emphasizing tokens based on attention-derived dependencies rather than random masking.
Prompt template choice strongly affects apparent performance of parallel decoding methods in dLLMs, causing inconsistent rankings and illusory gains; current methods fail to beat single-token decoding.
MRP predicts logit residuals between adjacent denoising steps in DLMs from backbone hidden states to support efficient multi-token denoising, yielding up to 1.4x lossless speedup or 22.6-point accuracy gains on code and math tasks.
citing papers explorer
-
Unified Energy for Invariant and Independent Decoding in Diffusion Language Models
The paper introduces Uni-E, a unified energy for DLMs that accounts for model capacity, dependency and invariance, can be computed exactly, and corrects distribution shifts from dependency and invariance.
-
Factorization-Error-Free Discrete Diffusion Language Model via Speculative Decoding
FeF-DLLM achieves factorization-error-free generation in discrete diffusion language models via prefix-conditioned posterior factorization and speculative decoding, delivering 5.04 pp higher accuracy and 3.86x faster inference on GSM8K, MATH, HumanEval, and MBPP.
-
Dynamic Execution Commitment of Vision-Language-Action Models
A3 reframes dynamic action chunk commitment in VLA models as self-speculative prefix verification, accepting the longest continuous sequence of actions that satisfies consensus-ordered conditional invariance and prefix-closed sequential consistency.
-
MARS: Enabling Autoregressive Models Multi-Token Generation
MARS fine-tunes autoregressive models to predict multiple tokens per step via continued training on instruction data, achieving 1.5-1.7x throughput while matching baseline accuracy and supporting real-time speed adjustment.
-
TEAM: Temporal-Spatial Consistency Guided Expert Activation for MoE Diffusion Language Model Acceleration
TEAM accelerates MoE dLLMs up to 2.2x by exploiting temporal-spatial consistency in expert routing to accept more tokens with fewer activations.
-
Accelerating Speculative Diffusions via Block Verification
A new residual-sampling scheme for diffusion models permits block verification and yields up to 6.3% speedup via a heuristic self-speculative drafter that needs no training.
-
Beyond Fully Random Masking: Attention-Guided Denoising and Optimization for Diffusion Language Models
AGDO improves dLLM reasoning performance by determining denoising order and emphasizing tokens based on attention-derived dependencies rather than random masking.
-
Understanding Evaluation Illusion in Diffusion Large Language Models
Prompt template choice strongly affects apparent performance of parallel decoding methods in dLLMs, causing inconsistent rankings and illusory gains; current methods fail to beat single-token decoding.
-
Multi-Token Residual Prediction
MRP predicts logit residuals between adjacent denoising steps in DLMs from backbone hidden states to support efficient multi-token denoising, yielding up to 1.4x lossless speedup or 22.6-point accuracy gains on code and math tasks.