[
    {
        "id": "osp-23037",
        "type": "article-journal",
        "title": "Does Uniform Discrete Diffusion Need Time?",
        "author": [
            {
                "family": "Hong",
                "given": "Chunsan"
            },
            {
                "family": "Lai",
                "given": "Chieh-Hsin"
            },
            {
                "family": "Hayakawa",
                "given": "Satoshi"
            },
            {
                "family": "Takida",
                "given": "Yuhta"
            },
            {
                "family": "Ye",
                "given": "Jong Chul"
            },
            {
                "family": "Mitsufuji",
                "given": "Yuki"
            }
        ],
        "URL": "https://omanscience.com/ar/articles/does-uniform-discrete-diffusion-need-time",
        "language": "en",
        "issued": {
            "date-parts": [
                [
                    2026
                ]
            ]
        },
        "abstract": "Uniform discrete diffusion models (UDMs) commonly use explicit time conditioning, but we find that it can often be unnecessary in practice. In this paper, we first show that the population-optimal UDM predictor generally depends on time: time controls how much the model should trust the observed context. We then show that this dependence can become negligible in finite-data settings relevant to language. When a corrupted training sequence remains much closer to its original clean sequence than to competing training sequences, the empirical-optimal predictor is nearly insensitive to time over most of the diffusion trajectory, where the guarantee weakens toward the high-noise endpoint. Empirically, trained language UDMs exhibit limited time sensitivity over most of the trajectory, while time-agnostic predictors remain competitive with, and often outperform, time-conditioned models across datasets and training objectives. These results challenge the use of explicit time conditioning in UDMs: although the population optimum depends on time, explicitly conditioning on it may often be unnecessary in practice."
    }
]