diffusion-experiments/index.html at main · bayesflow-org/diffusion-experiments · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">

  <!-- Primary Meta Tags -->
  <meta name="title" content="Diffusion Models In Simulation-Based Inference: A Tutorial Review - Arruda et al.">
  <meta name="description" content="Diffusion models have recently emerged as powerful learners for simulation-based inference (SBI), enabling fast and accurate estimation of latent parameters from simulated and real data.
Their score-based formulation offers a flexible way to learn conditional or joint distributions over parameters and observations, thereby providing a versatile solution to various modeling problems.
In this tutorial review, we synthesize recent developments on diffusion models for SBI, covering design choices for training, inference, and evaluation.
We highlight opportunities created by various concepts such as guidance, score composition, flow matching, consistency models, and joint modeling.
Furthermore, we discuss how efficiency and statistical accuracy are affected by noise schedules, parameterizations, and samplers.
Finally, we illustrate these concepts with case studies across parameter dimensionalities, simulation budgets, and model types, and outline open questions for future research.">
  <meta name="keywords" content="Simulation-Based Inference, Diffusion Models, Bayesian Inference, Statistical Methods">
  <meta name="author" content="Jonas Arruda, Niels Bracher, Ullrich Köthe, Jan Hasenauer, Stefan T. Radev">
  <meta name="robots" content="index, follow">
  <meta name="language" content="English">

  <!-- Open Graph / Facebook -->
  <meta property="og:type" content="article">
  <meta property="og:site_name" content="University of Bonn">
  <meta property="og:title" content="Diffusion Models In Simulation-Based Inference: A Tutorial Review">
  <!-- Same as description above -->
  <meta property="og:description" content="Diffusion models have recently emerged as powerful learners for simulation-based inference (SBI), enabling fast and accurate estimation of latent parameters from simulated and real data.
Their score-based formulation offers a flexible way to learn conditional or joint distributions over parameters and observations, thereby providing a versatile solution to various modeling problems.
In this tutorial review, we synthesize recent developments on diffusion models for SBI, covering design choices for training, inference, and evaluation.
We highlight opportunities created by various concepts such as guidance, score composition, flow matching, consistency models, and joint modeling.
Furthermore, we discuss how efficiency and statistical accuracy are affected by noise schedules, parameterizations, and samplers.
Finally, we illustrate these concepts with case studies across parameter dimensionalities, simulation budgets, and model types, and outline open questions for future research.">
  <meta property="og:url" content="https://bayesflow-org.github.io/diffusion-experiments/">
  <meta property="og:image" content="https://bayesflow-org.github.io/diffusion-experiments/static/images/social_preview.png">
  <meta property="og:image:width" content="1200">
  <meta property="og:image:height" content="630">
  <meta property="og:image:alt" content="Diffusion Models In Simulation-Based Inference: A Tutorial Review">
  <meta property="article:published_time" content="2025-12-22T00:00:00.000Z">
  <meta property="article:author" content="Arruda, Jonas; Bracher, Niels; Köthe, Ullrich; Hasenauer, Jan; Radev, Stefan T.">
  <meta property="article:section" content="Research">
  <meta property="article:tag" content="Simulation-Based Inference">
  <meta property="article:tag" content="Diffusion Models">


  <!-- Academic/Research Specific -->
  <meta name="citation_title" content="Diffusion Models In Simulation-Based Inference: A Tutorial Review">
  <meta name="citation_author" content="Arruda, Jonas">
  <meta name="citation_author" content="Bracher, Niels">
  <meta name="citation_author" content="Köthe, Ullrich">
  <meta name="citation_author" content="Hasenauer, Jan">
  <meta name="citation_author" content="Radev, Stefan T.">
  <meta name="citation_publication_date" content="2025">
  <meta name="citation_conference_title" content="arXiv preprint arXiv:2512.20685">
  <meta name="citation_pdf_url" content="https://arxiv.org/pdf/2512.20685">

  <!-- Additional SEO -->
  <meta name="theme-color" content="#2563eb">
  <meta name="msapplication-TileColor" content="#2563eb">
  <meta name="apple-mobile-web-app-capable" content="yes">
  <meta name="apple-mobile-web-app-status-bar-style" content="default">

  <!-- Preconnect for performance (local only) -->
  <link rel="preconnect" href="/">

  <title>Diffusion Models In Simulation-Based Inference: A Tutorial Review - Arruda et al.</title>

  <!-- Critical CSS - Load synchronously -->
  <link rel="stylesheet" href="static/css/bulma.min.css">
  <link rel="stylesheet" href="static/css/index.css">

  <!-- Non-critical CSS - Load asynchronously -->
  <link rel="preload" href="static/css/bulma-carousel.min.css" as="style" onload="this.onload=null;this.rel='stylesheet'">
  <link rel="preload" href="static/css/bulma-slider.min.css" as="style" onload="this.onload=null;this.rel='stylesheet'">
  <link rel="preload" href="static/css/fontawesome.all.min.css" as="style" onload="this.onload=null;this.rel='stylesheet'">

  <!-- Fallback for browsers that don't support preload -->
  <noscript>
    <link rel="stylesheet" href="static/css/bulma-carousel.min.css">
    <link rel="stylesheet" href="static/css/bulma-slider.min.css">
    <link rel="stylesheet" href="static/css/fontawesome.all.min.css">
  </noscript>

  <!-- Defer non-critical JavaScript -->
  <script defer src="static/js/fontawesome.all.min.js"></script>
  <script defer src="static/js/bulma-carousel.min.js"></script>
  <script defer src="static/js/bulma-slider.min.js"></script>
  <script defer src="static/js/index.js"></script>
  <!-- Structured Data for Academic Papers -->
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "ScholarlyArticle",
  "headline": "Diffusion Models In Simulation-Based Inference: A Tutorial Review",
  "description": "A comprehensive tutorial review of diffusion models for simulation-based inference, covering theoretical foundations, design choices, efficiency–accuracy trade-offs, and practical case studies across diverse inference settings.",
  "author": [
    {
      "@type": "Person",
      "name": "Jonas Arruda",
      "affiliation": {
        "@type": "Organization",
        "name": "University of Bonn"
      }
    },
    {
      "@type": "Person",
      "name": "Niels Bracher",
      "affiliation": {
        "@type": "Organization",
        "name": "Rensselaer Polytechnic Institute"
      }
    },
    {
      "@type": "Person",
      "name": "Ullrich Köthe",
      "affiliation": {
        "@type": "Organization",
        "name": "Heidelberg University"
      }
    },
    {
      "@type": "Person",
      "name": "Jan Hasenauer",
      "affiliation": {
        "@type": "Organization",
        "name": "University of Bonn"
      }
    },
    {
      "@type": "Person",
      "name": "Stefan T. Radev",
      "affiliation": {
        "@type": "Organization",
        "name": "Rensselaer Polytechnic Institute"
      }
    }
  ],
  "datePublished": "2025-12-22",
  "publisher": {
    "@type": "Organization",
    "name": "arXiv"
  },
  "url": "https://bayesflow-org.github.io/diffusion-experiments/",
  "image": "https://bayesflow-org.github.io/diffusion-experiments/static/images/social_preview.png",
  "keywords": [
    "simulation-based inference",
    "diffusion models",
    "Bayesian inference",
    "score-based models",
    "probabilistic machine learning"
  ],
  "abstract": "Diffusion models have recently emerged as powerful learners for simulation-based inference (SBI), enabling fast and accurate estimation of latent parameters from simulated and real data. Their score-based formulation offers a flexible way to learn conditional or joint distributions over parameters and observations, thereby providing a versatile solution to various modeling problems. In this tutorial review, we synthesize recent developments on diffusion models for SBI, covering design choices for training, inference, and evaluation. We highlight opportunities created by various concepts such as guidance, score composition, flow matching, consistency models, and joint modeling. Furthermore, we discuss how efficiency and statistical accuracy are affected by noise schedules, parameterizations, and samplers. Finally, we illustrate these concepts with case studies across parameter dimensionalities, simulation budgets, and model types, and outline open questions for future research.",
  "citation": "@article{arruda2025diffusionSBI, title={Diffusion Models in Simulation-Based Inference: A Tutorial Review}, author={Arruda, Jonas and Bracher, Niels and Köthe, Ullrich and Hasenauer, Jan and Radev, Stefan T}, journal={arXiv preprint arXiv:2512.20685}, year={2025}, url={https://doi.org/10.48550/arXiv.2512.20685} }",
  "isAccessibleForFree": true,
  "license": "https://creativecommons.org/licenses/by/4.0/",
  "mainEntity": {
    "@type": "WebPage",
    "@id": "https://bayesflow-org.github.io/diffusion-experiments/"
  },
  "about": [
    {
      "@type": "Thing",
      "name": "Simulation-Based Inference"
    },
    {
      "@type": "Thing",
      "name": "Diffusion Models"
    },
    {
      "@type": "Thing",
      "name": "Statistics"
    }
  ]
}
</script>

</head>
<body>


  <!-- Scroll to Top Button -->
  <button class="scroll-to-top" onclick="scrollToTop()" title="Scroll to top" aria-label="Scroll to top">
    <i class="fas fa-chevron-up"></i>
  </button>

  <main id="main-content">
  <section class="hero">
    <div class="hero-body">
      <div class="container is-max-desktop">
        <div class="columns is-centered">
          <div class="column has-text-centered">
            <h1 class="title is-1 publication-title">Diffusion Models In Simulation-Based Inference: A Tutorial Review</h1>
            <div class="is-size-5 publication-authors">
              <span class="author-block">
                <a href="https://github.com/arrjon" target="_blank">Jonas Arruda</a>,</span>
                <span class="author-block">
                  <a href="https://bayesops.com/members/niels-bracher" target="_blank">Niels Bracher</a>,</span>
              <span class="author-block">
                  <a href="https://scholar.google.com/citations?user=gt-yaNMAAAAJ&hl=de" target="_blank">Ullrich Köthe</a>,</span>
              <span class="author-block">
                  <a href="https://www.mathematics-and-life-sciences.uni-bonn.de/en/group-members/people/hasenauer-group-members/jan-hasenauer" target="_blank">Jan Hasenauer</a>,</span>
                <span class="author-block">
                    <a href="https://bayesops.com/members/stefan-radev.html" target="_blank">Stefan T. Radev</a>
                </span>
            </div>

                  <div class="is-size-5 publication-authors">
                    <span class="author-block">University of Bonn<br>Rensselaer Polytechnic Institute<br>Heidelberg University</span>
                  </div>

                  <div class="column has-text-centered">
                    <div class="publication-links">
                      <span class="link-block">
                        <a href="https://arxiv.org/pdf/2512.20685" target="_blank"
                        class="external-link button is-normal is-rounded is-dark">
                        <span class="icon">
                          <i class="fas fa-file-pdf"></i>
                        </span>
                        <span>Paper</span>
                      </a>
                    </span>

                  <span class="link-block">
                    <a href="https://github.com/bayesflow-org/diffusion-experiments" target="_blank"
                    class="external-link button is-normal is-rounded is-dark">
                    <span class="icon">
                      <i class="fab fa-github"></i>
                    </span>
                    <span>Code</span>
                  </a>
                </span>

                <span class="link-block">
                  <a href="https://arxiv.org/abs/2512.20685" target="_blank"
                  class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                   <i class="fas fa-scroll"></i>
                   </span>
                  <span>arXiv</span>
                </a>
              </span>
                      <span class="link-block">
                        <a href="https://youtu.be/ZlcEkHXgF5k" target="_blank"
                        class="external-link button is-normal is-rounded is-dark">
                        <span class="icon">
                          <i class="fas fa-video"></i>
                        </span>
                        <span>Tutorial</span>
                      </a>
                    </span>
            </div>
          </div>
        </div>
      </div>
    </div>
  </div>
</section>


<!-- Paper abstract -->
<section class="section hero is-light">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3">Abstract</h2>
        <div class="content has-text-justified">
          <p>
            Diffusion models have recently emerged as powerful learners for simulation-based inference (SBI), enabling fast and accurate estimation of latent parameters from simulated and real data.
Their score-based formulation offers a flexible way to learn conditional or joint distributions over parameters and observations, thereby providing a versatile solution to various modeling problems.
In this tutorial review, we synthesize recent developments on diffusion models for SBI, covering design choices for training, inference, and evaluation.
We highlight opportunities created by various concepts such as guidance, score composition, flow matching, consistency models, and joint modeling.
Furthermore, we discuss how efficiency and statistical accuracy are affected by noise schedules, parameterizations, and samplers.
Finally, we illustrate these concepts with case studies across parameter dimensionalities, simulation budgets, and model types, and outline open questions for future research.
          </p>
        </div>
      </div>
    </div>
  </div>
</section>
<!-- End paper abstract -->


<!-- Image carousel -->
<section class="hero is-small">
  <div class="hero-body">
    <div class="container is-max-desktop">
      <div id="results-carousel" class="carousel results-carousel">
       <div class="item">
        <img src="static/images/diffusion_model_review.jpg" alt=" What is SBI? What are diffusion models?" loading="lazy"/>
        <h2 class="subtitle has-text-centered">
          What is SBI? What are diffusion models?
        </h2>
      </div>
      <div class="item">
        <!-- Your image here -->
        <img src="static/images/adaptive_inference.jpg" alt="What makes diffusion models special in SBI?" loading="lazy"/>
        <h2 class="subtitle has-text-centered">
          What is special about diffusion models in SBI?
        </h2>
      </div>
      <div class="item">
        <!-- Your image here -->
        <img src="static/images/c2st_benchmark_boxplot_best.jpg" alt="Which are the key design choices? Which design for which problem?" loading="lazy"/>
        <h2 class="subtitle has-text-centered">
         Which are the key design choices? Which design for which problem?
       </h2>
     </div>
  </div>
</div>
</div>
</section>
<!-- End image carousel -->
<section class="section">
    <div class="container is-max-desktop">

      <div id="marimo-wrapper" style="border: 1px solid #ddd; border-radius: 8px; padding: 1rem;">

        <button id="load-marimo" type="button"
          style="padding: .5rem .75rem; border: 1px solid #ccc; border-radius: 6px; background: white; cursor: pointer;">
          <strong>Show me the tutorial notebook</strong>
        </button>

        <noscript>
          <p style="margin-top: 0.75rem;">
            JavaScript is required to display the notebook. You can open it directly:
            <a href="static/tutorial_plain.html">static/tutorial_plain.html</a>
          </p>
        </noscript>
      </div>

      <div id="marimo-frame-slot" style="margin-top: 1rem;"></div>

    </div>
</section>

<script>
  (function () {
    const btn = document.getElementById("load-marimo");
    const slot = document.getElementById("marimo-frame-slot");
    const wrapper = document.getElementById("marimo-wrapper");

    btn.addEventListener("click", function () {
      // prevent double insertion
      if (slot.firstChild) return;

      const iframe = document.createElement("iframe");
      iframe.src = "static/tutorial_plain.html";
      iframe.style.width = "100%";
      iframe.style.height = "100vh";
      iframe.style.border = "none";

      // privacy / containment
      iframe.setAttribute("referrerpolicy", "no-referrer");
      iframe.setAttribute("sandbox", "allow-scripts allow-same-origin allow-popups");

      slot.appendChild(iframe);

      // optional: replace the hint after loading
      wrapper.innerHTML =
        '<p style="margin:0;"><strong>Tutorial</strong></p>';
    });
  })();
</script>

<section class="section">
    <div class="container is-max-desktop">

      <div class="is-flex is-justify-content-center">
        <div style="max-width: 640px; width: 100%;">

          <a href="https://youtu.be/ZlcEkHXgF5k"
             target="_blank"
             rel="noopener noreferrer"
             style="display: block; border: 1px solid #ddd; border-radius: 8px; overflow: hidden; text-decoration: none; color: inherit;">

            <img src="static/images/youtube_podcast_tumbnail.png"
                 alt="Watch the tutorial podcast on diffusion models in SBI on YouTube"
                 style="display: block; width: 100%; height: auto;">

            <div style="padding: 0.75rem; background: #fff; text-align: center;">
              <strong>Watch the tutorial podcast on diffusion models in SBI on YouTube</strong>
            </div>

          </a>

        </div>
      </div>

    </div>
</section>


<!--BibTex citation -->
  <section class="section" id="BibTeX">
    <div class="container is-max-desktop content">
      <div class="bibtex-header">
        <h2 class="title">BibTeX</h2>
        <button class="copy-bibtex-btn" onclick="copyBibTeX()" title="Copy BibTeX to clipboard">
          <i class="fas fa-copy"></i>
          <span class="copy-text">Copy</span>
        </button>
      </div>
      <pre id="bibtex-code"><code>@article{arruda2025diffusionSBI,
  title={Diffusion Models in Simulation-Based Inference: A Tutorial Review},
  author={Arruda, Jonas and Bracher, Niels and K{\"o}the, Ullrich and Hasenauer, Jan and Radev, Stefan T},
  journal={arXiv preprint arXiv:2512.20685},
  year={2025},
  url={https://doi.org/10.48550/arXiv.2512.20685}
}</code></pre>
    </div>
</section>
<!--End BibTex citation -->


  <footer class="footer">
  <div class="container">
    <div class="columns is-centered">
      <div class="column is-8">
        <div class="content">

          <p>
            This page was built using the <a href="https://github.com/eliahuhorwitz/Academic-project-page-template" target="_blank">Academic Project Page Template</a> which was adopted from the <a href="https://nerfies.github.io" target="_blank">Nerfies</a> project page.
            Commons Attribution-ShareAlike 4.0 International License</a>.
          </p>

        </div>
      </div>
    </div>

  <div class="content has-text-centered is-size-7">
    <details>
      <summary style="cursor:pointer;">Impressum</summary>
       <p>
        This website is hosted via GitHub Pages (GitHub, Inc.). No tracking, cookies, or analytics are used.
        GitHub may process technical access data (e.g., IP address and browser information) to deliver the site.
      </p>

      <p>Jonas Arruda · University of Bonn · Endenicher Allee 64 · 53115 Bonn</p>
      <p><span>jonas.arruda</span> [at] <span>uni-bonn.de</span></p>
    </details>
  </div>

  </div>
</footer>

</body>
</html>