# Search or Accelerate: Confidence-Switched Position Beam Search for Diffusion Language Models

arXiv: https://arxiv.org/abs/2602.10953 · HTML: https://mlsanity.com/paper/2602.10953/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Discrete and text diffusion, one of 14 directions in Diffusion models.
- Builds on 22 of its 25 references that are in the map.
- First posted on arXiv 11 Feb 2026.

## Placed in

- Diffusion models / [Discrete and text diffusion](https://mlsanity.com/diffusion/discrete-and-text-diffusion/)

## Builds on (most cited first, 22 of 22)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Dream 7B: Diffusion Large Language Models](https://mlsanity.com/paper/2508.15487/) (2025)
- [Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding](https://mlsanity.com/paper/2505.22618/) (2025)
- [LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models](https://mlsanity.com/paper/2505.19223/) (2025)
- [Scaling Diffusion Language Models via Adaptation from Autoregressive Models](https://mlsanity.com/paper/2410.17891/) (2024)
- [Efficient Training of Language Models to Fill in the Middle](https://mlsanity.com/paper/2207.14255/) (2022)
- [Beam Search Strategies for Neural Machine Translation](https://mlsanity.com/paper/1702.01806/) (2017)
- [DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation](https://mlsanity.com/paper/2506.20639/) (2025)
- [dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Caching](https://mlsanity.com/paper/2506.06295/) (2025)
- [dKV-Cache: The Cache for Diffusion Language Models](https://mlsanity.com/paper/2505.15781/) (2025)
- [Diffusion Language Models Know the Answer Before Decoding](https://mlsanity.com/paper/2508.19982/) (2025)
- [Don't Settle Too Early: Self-Reflective Remasking for Diffusion Language Models](https://mlsanity.com/paper/2509.23653/) (2025)
- [Beyond Fixed: Training-Free Variable-Length Denoising for Diffusion Large Language Models](https://mlsanity.com/paper/2508.00819/) (2025)
- [Attention Is All You Need for KV Cache in Diffusion LLMs](https://mlsanity.com/paper/2510.14973/) (2025)
- [Self Speculative Decoding for Diffusion Large Language Models](https://mlsanity.com/paper/2510.04147/) (2025)
- [Diffusion LLM with Native Variable Generation Lengths: Let [EOS] Lead the Way](https://mlsanity.com/paper/2510.24605/) (2025)
- [DreamOn: Diffusion Language Models For Code Infilling Beyond Fixed-size Canvas](https://mlsanity.com/paper/2602.01326/) (2026)
- [Mask Tokens as Prophet: Fine-Grained Cache Eviction for Efficient dLLM Inference](https://mlsanity.com/paper/2510.09309/) (2025)
- [Improving Diffusion Language Model Decoding through Joint Search in Generation Order and Token Space](https://mlsanity.com/paper/2601.20339/) (2026)
- [Latent Refinement Decoding: Enhancing Diffusion-Based Language Models by Refining Belief States](https://mlsanity.com/paper/2510.11052/) (2025)

## Built on by (most cited first, 3 of 3)

- [Focus on the Core: Empowering Diffusion Large Language Models by Self-Contrast](https://mlsanity.com/paper/2605.01373/) (2026)
- [Constraints Are Graphs, Not Chains: Exact Decoding for Diffusion Language Models](https://mlsanity.com/paper/2609.32900/) (2026)
- [Unmask the State: When Does State Adaptation Matter for Masked Diffusion Language Models](https://mlsanity.com/paper/2609.33355/) (2026)

## Abstract (arXiv)

Diffusion Language Models (DLMs) generate text by iteratively denoising a masked sequence, repeatedly deciding which positions to commit at each step. Standard decoding follows a greedy rule: unmask the most confident positions, yet this local choice can lock the model into a suboptimal unmasking order, especially on reasoning-heavy prompts. We present SOAR, a training-free decoding algorithm that adapts its behavior to the model's uncertainty. When confidence is low, SOAR briefly widens the search over alternative unmasking decisions to avoid premature commitments; when confidence is high, it collapses the search and decodes many positions in parallel to reduce the number of denoising iterations. Across mathematical reasoning and code generation benchmarks (GSM8K, MBPP, HumanEval) on Dream-7B and LLaDA-8B, SOAR improves generation quality while maintaining competitive inference speed, offering a practical way to balance quality and efficiency in DLM decoding. Our Code is available at https://github.com/duterscmy/SOAR