Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

What Gets Unmasked First? Trajectory Analysis of Diffusion Models for Graph-to-Text Generation

About

We present the first systematic study of masked diffusion language models (MDLMs) for graph-to-text generation. We analyze MDLM generation trajectories -- the order in which tokens are unmasked during iterative decoding -- and find that, unlike autoregressive LLMs which generate text linearly, MDLMs naturally prioritize entities first, followed by relational and function words, with structural tokens resolved last. We further identify a previously undocumented failure mode of supervised fine-tuning: SFT disrupts this strategy by prematurely anchoring structural sentence-ending tokens early in the decoding trajectory, effectively fixing the output length which can lead to omitted or hallucinated information. To address this, we propose lambda-scaled structural decoding, a training-free inference-time modification that downweights structural token confidence and recovers +9.4 BLEU-4. Finally, we introduce Graph-LLaDA, which integrates a Graph Transformer encoder into LLaDA's decoding process to explicitly incorporate relational graph structure. Cross-dataset evaluation on LAGRANGE reveals that previous baselines overfit to dataset-specific patterns, while LLM- and MDLM-based approaches generalize significantly better.

Qing Wang, Jacob Devasier, Chengkai Li• 2026

Related benchmarks

TaskDatasetResultRank
Graph-to-text generationWebNLG English (test)
BLEU-455.4
14
Graph-to-TextLAGRANGE (test)
Avg Elo1.60e+3
7
Graph-to-text generationGraph-LLaDA-generated sentences
Fluency100
4
Showing 3 of 3 rows

Other info

Follow for update