# Simple Self-Conditioning Adaptation for Masked Diffusion Models

arXiv: https://arxiv.org/abs/2604.26985 · HTML: https://mlsanity.com/paper/2604.26985/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Discrete and text diffusion, one of 14 directions in Diffusion models.
- Builds on 21 of its 30 references that are in the map.
- First posted on arXiv 28 Apr 2026.

## Placed in

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

## Builds on (most cited first, 21 of 21)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Structured Denoising Diffusion Models in Discrete State-Spaces](https://mlsanity.com/paper/2107.03006/) (2021)
- [Large Language Diffusion Models](https://mlsanity.com/paper/2502.09992/) (2025)
- [Simple and Effective Masked Diffusion Language Models](https://mlsanity.com/paper/2406.07524/) (2024)
- [Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution](https://mlsanity.com/paper/2310.16834/) (2023)
- [Dream 7B: Diffusion Large Language Models](https://mlsanity.com/paper/2508.15487/) (2025)
- [Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning](https://mlsanity.com/paper/2208.04202/) (2022)
- [A Survey on Generative Diffusion Model](https://mlsanity.com/paper/2209.02646/) (2022)
- [Mercury: Ultra-Fast Language Models Based on Diffusion](https://mlsanity.com/paper/2506.17298/) (2025)
- [Masked Diffusion Models are Secretly Time-Agnostic Masked Models and Exploit Inaccurate Categorical Sampling](https://mlsanity.com/paper/2409.02908/) (2024)
- [DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation](https://mlsanity.com/paper/2506.20639/) (2025)
- [Simple Guidance Mechanisms for Discrete Diffusion Models](https://mlsanity.com/paper/2412.10193/) (2024)
- [Concrete Score Matching: Generalized Score Matching for Discrete Data](https://mlsanity.com/paper/2211.00802/) (2022)
- [Self-conditioned Embedding Diffusion for Text Generation](https://mlsanity.com/paper/2211.04236/) (2022)
- [Discrete Diffusion in Large Language and Multimodal Models: A Survey](https://mlsanity.com/paper/2506.13759/) (2025)
- [Speculative Diffusion Decoding: Accelerating Language Generation through Diffusion](https://mlsanity.com/paper/2408.05636/) (2024)
- [Constrained Discrete Diffusion](https://mlsanity.com/paper/2503.09790/) (2025)
- [Residual Context Diffusion Language Models](https://mlsanity.com/paper/2601.22954/) (2026)
- [Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall](https://mlsanity.com/paper/2510.19304/) (2025)
- [Search-Augmented Masked Diffusion Models for Constrained Generation](https://mlsanity.com/paper/2602.02727/) (2026)

## Built on by (most cited first, 4 of 4)

- [Neither Parallel Nor Sequential: How DiffusionGemma Actually Commits Tokens](https://mlsanity.com/paper/2606.14620/) (2026)
- [Improving Few-Step Language Flows with Untied Self-Conditioning](https://mlsanity.com/paper/2608.22244/) (2026)
- [ALoDLM: Adaptively Looped Diffusion Language Models](https://mlsanity.com/paper/2610.04198/) (2026)
- [Only What You've Signed up For: Secure Workflows with Constrained Diffusion Agents](https://mlsanity.com/paper/2610.15177/) (2026)

## Abstract (arXiv)

Masked diffusion models (MDMs) generate discrete sequences by iterative denoising under an absorbing masking process. In standard masked diffusion, if a token remains masked after a reverse update, the model discards its clean-state prediction for that position. Thus, still-masked positions must be repeatedly inferred from the mask token alone. This design choice limits cross-step refinement. To address this limitation, this paper proposes a simple, yet effective, post-training adaptation for MDMs that conditions each denoising step on the model's own previous clean-state predictions. The resulting method, called Self-Conditioned Masked Diffusion Models (SCMDM), requires minimal architectural change, does not introduce a recurrent latent-state pathway, does not rely on an auxiliary reference model, and adds no extra denoiser evaluations during sampling. This is an important departure from partial self-conditioning approaches which requires expensive model training from scratch. In particular, the paper shows that partial self-conditioning, including the commonly used 50% dropout strategy for training self-conditioned models from scratch, is suboptimal in the post-training regime. Instead, once the model's self-generated clean-state estimates become informative, the specialization to refinement is preferable to mixing conditional and unconditional objectives. SCMDM is evaluated across multiple domains, demonstrating consistent improvement over vanilla MDM baselines, achieving nearly a 50% reduction in generative perplexity on OWT-trained models (42.89 to 23.72), alongside strong improvements in discretized image synthesis quality, small molecular generation, and enhanced fidelity in genomic distribution modeling.