Can I Have Your Order? Monte-Carlo Tree Search for Slot Filling Ordering in Diffusion Language Models

TL;DR

This paper introduces McDiffuSE, a Monte Carlo Tree Search framework that optimizes slot filling order in Masked Diffusion Models, significantly improving reasoning task performance by systematic exploration of generation sequences.

Contribution

It presents a novel MCTS-based method for planning slot infilling order in diffusion models, outperforming existing autoregressive and plan-and-infill baselines.

Findings

Average performance improvement of 3.2% over autoregressive methods.

Achieves 8.0% better results than baseline plan-and-infill.

Notable gains of 19.5% on MBPP and 4.9% on MATH500.

Abstract

While plan-and-infill decoding in Masked Diffusion Models (MDMs) shows promise for mathematical and code reasoning, performance remains highly sensitive to slot infilling order, often yielding substantial output variance. We introduce McDiffuSE, a framework that formulates slot selection as decision making and optimises infilling orders through Monte Carlo Tree Search (MCTS). McDiffuSE uses look-ahead simulations to evaluate partial completions before commitment, systematically exploring the combinatorial space of generation orders. Experiments show an average improvement of 3.2% over autoregressive baselines and 8.0% over baseline plan-and-infill, with notable gains of 19.5% on MBPP and 4.9% on MATH500. Our analysis reveals that while McDiffuSE predominantly follows sequential ordering, incorporating non-sequential generation is essential for maximising performance. We observe that larger exploration constants, rather than increased simulations, are necessary to overcome model confidence biases and discover effective orderings. These findings establish MCTS-based planning as an effective approach for enhancing generation quality in MDMs.