implicitvalue.github.io/index.html at main · implicitvalue/implicitvalue.github.io · 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
<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1" />
  <title>Yunzhen Yao</title>
  <meta name="description" content="Yunzhen Yao — PhD student at EPFL." />

  <!--
    Template style goals:
    - Minimal, content-first layout
    - Single-page academic content blocks
    - No build system required: works as a standalone index.html for GitHub Pages
  -->

  <style>
    :root{
      --bg:#ffffff;
      --fg:#111111;
      --muted:#666666;
      --link:#1a73e8;
      --border:#e8e8e8;
      --card:#fafafa;
      --code:#f5f5f5;
      --maxw: 860px;
    }
    @media (prefers-color-scheme: dark){
      :root{
        --bg:#0b0f14;
        --fg:#eaecef;
        --muted:#a8b3bf;
        --link:#7ab7ff;
        --border:#1f2a35;
        --card:#0f1620;
        --code:#111b27;
      }
    }

    html, body { height: 100%; }
    body{
      margin:0;
      background:var(--bg);
      color:var(--fg);
      font-family: -apple-system,BlinkMacSystemFont,"Segoe UI",Roboto,Helvetica,Arial,"Noto Sans",sans-serif;
      line-height:1.65;
      -webkit-font-smoothing: antialiased;
      -moz-osx-font-smoothing: grayscale;
    }

    a{ color:var(--link); text-decoration:none; }
    a:hover{ text-decoration:underline; }

    .wrap{
      max-width: var(--maxw);
      margin: 0 auto;
      padding: 28px 18px 72px;
    }

    /* Top nav */
    .topbar{
      display:flex;
      align-items:center;
      justify-content:space-between;
      gap:14px;
      flex-wrap:wrap;
      padding-bottom:14px;
      border-bottom:1px solid var(--border);
      margin-bottom:22px;
    }
    .site-title{
      font-weight: 800;
      letter-spacing: -0.02em;
      font-size: 1.05rem;
      margin:0;
      white-space: nowrap;
    }
    .nav{
      display:flex;
      gap:14px;
      flex-wrap:wrap;
      align-items:center;
      font-size: 0.98rem;
    }
    .nav a{
      font-weight:600;
      color: var(--fg);
    }
    .nav a:hover{ color: var(--link); text-decoration:none; }

    /* Header block */
    .hero{
      display:grid;
      grid-template-columns: 150px 1fr;
      gap:18px;
      align-items:start;
      margin-top: 8px;
    }
    @media (max-width: 560px){
      .hero{ grid-template-columns: 1fr; }
    }

    .avatar{
      width:120px;
      height:160px;
      border-radius: 14px;
      background: var(--card);
      border:1px solid var(--border);
      display:flex;
      align-items:center;
      justify-content:center;
      overflow:hidden;
    }
    .avatar img{
      width:100%;
      height:100%;
      object-fit:cover;
      display:block;
    }

    h1{
      margin: 0 0 6px 0;
      font-size: 2rem;
      line-height: 1.15;
      letter-spacing: -0.02em;
    }
    .subtitle{
      margin: 0 0 10px 0;
      color: var(--muted);
      font-size: 1.02rem;
    }
    .linkrow{
      display:flex;
      flex-wrap:wrap;
      gap:20px;
      margin: 8px 0 0 0;
      font-weight: 600;
    }
    .linkrow a{ color: var(--link); }

    .email{
      margin-top: 8px;
      color: var(--muted);
      font-size: 0.98rem;
    }

    section{ margin-top: 26px; }
    h2{
      font-size: 1.15rem;
      margin: 0 0 10px 0;
      letter-spacing: -0.01em;
    }

    .card{
      background: var(--card);
      border: 1px solid var(--border);
      border-radius: 14px;
      padding: 14px 16px;
    }
    .research{
      margin-top: 14px;
    }
    .research p{
      margin: 0;
    }
    .research p + p{
      margin-top: 6px;
    }
    .research ul{
      list-style-type: circle;
    }

    ul{ margin: 10px 0 0 0; padding-left: 18px; }
    li{ margin: 7px 0; }

    .news li{ color: var(--fg); }
    .news .date{
      color: var(--muted);
      font-variant-numeric: tabular-nums;
      margin-right: 8px;
      white-space: nowrap;
    }

    .pub{ margin: 0; padding-left: 18px; }
    .pub li{ margin: 12px 0; }
    .pub-title{ font-weight: 700; }
    .pub-meta{ color: var(--muted); }
    .pub-links{ margin-top: 3px; font-weight: 600; }
    .pub-links a{ margin-right: 8px; }

    .small{ color: var(--muted); font-size: 0.95rem; }

    code.inline{
      background: var(--code);
      border: 1px solid var(--border);
      padding: 1px 6px;
      border-radius: 10px;
      font-size: 0.95em;
    }

    footer{
      margin-top: 42px;
      padding-top: 16px;
      border-top: 1px solid var(--border);
      color: var(--muted);
      font-size: 0.92rem;
    }
  </style>

  <!-- Google tag (gtag.js) -->
  <script async src="https://www.googletagmanager.com/gtag/js?id=G-GX8QSZTJH6"></script>
  <script>
    window.dataLayer = window.dataLayer || [];
    function gtag(){dataLayer.push(arguments);}
    gtag('js', new Date());
    gtag('config', 'G-GX8QSZTJH6');
  </script>
</head>

<body>
  <div class="wrap">
    <!-- Header / Hero -->
    <div class="hero" id="about">
      <div class="avatar" aria-label="Profile photo">
        <img src="avatar.jpg" alt="Yunzhen Yao" />
      </div>

      <div>
        <h1>Yunzhen Yao</h1>
        <p class="subtitle">
          PhD Student @ EPFL · Machine Learning
        </p>

        <div class="linkrow">
          <a href="https://scholar.google.com/citations?user=Ga00jcIAAAAJ&hl=en"
             target="_blank" rel="noopener"
             data-ga="scholar" data-title="Google Scholar">
            Google Scholar
          </a>

          <a href="https://github.com/implicitvalue"
             target="_blank" rel="noopener"
             data-ga="github" data-title="GitHub Profile">
            GitHub
          </a>

          <a href="https://www.linkedin.com/in/yunzhen-yao-18a87a347"
             target="_blank" rel="noopener"
             data-ga="linkedin" data-title="LinkedIn">
            LinkedIn
          </a>

          <a href="https://dblp.org/pid/284/0567.html"
             target="_blank" rel="noopener"
             data-ga="dblp" data-title="DBLP">
            DBLP
          </a>

          <a href="https://www.dropbox.com/scl/fi/srobtguq6ptsocl057tuo/Yunzhen_Yao_CV_2026.pdf?rlkey=wfj5h3yn9vqkvksdp98v81rmy&st=ho7oktpe&dl=0"
             target="_blank" rel="noopener"
             data-ga="cv" data-title="CV (Dropbox PDF)">
            CV
          </a>
        </div>

        <div class="email">
          Email: firstnamelastname(at)gmail(dot)com
        </div>

      </div>
    </div>

    <div>
      <section class="research">
        <p>
          I am a fourth-year PhD student at
          <a href="https://www.epfl.ch/en/"
             target="_blank" rel="noopener"
             data-ga="affiliation" data-title="EPFL">
            EPFL
          </a>, advised by
          Prof.
          <a href="https://ic-people.epfl.ch/~gastpar/"
             target="_blank" rel="noopener"
             data-ga="person" data-title="Michael Gastpar (advisor)">
            Michael Gastpar
          </a>.
        </p>

        <p>
          My research focuses on the <strong>theoretical foundations and practical methods</strong> for building
          <strong>reliable, trustworthy, and efficient</strong> machine learning systems, with an emphasis on
          <strong>uncertainty quantification</strong> and <strong>robust decision-making</strong>.
        </p>

        <ul>
          <li>Inference-time scaling for large language models (LLMs)</li>
          <li>Uncertainty quantification for machine learning systems, including conformal prediction</li>
          <li>Preference learning and LLM alignment</li>
          <li>Finite-sample and non-asymptotic analysis of learning algorithms</li>
        </ul>
      </section>
    </div>

    <!-- Publications -->
    <section id="publications">
      <h2>Publications</h2>

      <ul class="pub">
        <li>
          <div class="pub-title">
            <a href="https://arxiv.org/abs/2510.07093"
               target="_blank" rel="noopener"
               data-ga="publication" data-title="Non-Asymptotic Analysis of Efficiency in Conformalized Regression (arXiv)">
              Non-Asymptotic Analysis of Efficiency in Conformalized Regression
            </a>
          </div>
          <div class="pub-meta">
            <em>Yunzhen Yao</em>, Lie He, Michael Gastpar · ICLR 2026
          </div>
        </li>

        <li>
          <div class="pub-title">
            <a href="https://arxiv.org/abs/2506.01084"
               target="_blank" rel="noopener"
               data-ga="publication" data-title="zip2zip (arXiv)">
              zip2zip: Inference-Time Adaptive Vocabularies for Language Models via Token Compression
            </a>
          </div>
          <div class="pub-meta">
            Saibo Geng*, Nathan Ranchin*, <em>Yunzhen Yao</em>, Maxime Peyrard,
            Chris Wendler, Michael Gastpar, Robert West · NeurIPS 2025
          </div>
        </li>

        <li>
          <div class="pub-title">
            <a href="https://arxiv.org/abs/2501.18282"
               target="_blank" rel="noopener"
               data-ga="publication" data-title="Leveraging Sparsity for Sample-Efficient Preference Learning (arXiv)">
              Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective
            </a>
          </div>
          <div class="pub-meta">
            <em>Yunzhen Yao</em>, Lie He, Michael Gastpar · ICML 2025
          </div>
        </li>

        <li>
          <div class="pub-title">
            <a href="https://www.jmlr.org/papers/volume25/22-0816/22-0816.pdf"
               target="_blank" rel="noopener"
               data-ga="publication" data-title="Unlabeled Principal Component Analysis and Matrix Completion (JMLR PDF)">
              Unlabeled Principal Component Analysis and Matrix Completion
            </a>
          </div>
          <div class="pub-meta">
            <em>Yunzhen Yao</em>, Liangzu Peng, Manolis Tsakiris · JMLR 2024
          </div>
        </li>

        <li>
          <div class="pub-title">
            <a href="https://proceedings.neurips.cc/paper/2021/hash/ff8c1a3bd0c441439a0a081e560c85fc-Abstract.html"
               target="_blank" rel="noopener"
               data-ga="publication" data-title="Unlabeled Principal Component Analysis (NeurIPS 2021)">
              Unlabeled Principal Component Analysis
            </a>
          </div>
          <div class="pub-meta">
            <em>Yunzhen Yao</em>, Liangzu Peng, Manolis Tsakiris · NeurIPS 2021
          </div>
        </li>

        <li>
          <div class="pub-title">
            <a href="https://ieeexplore.ieee.org/document/9517881/authors#authors"
               target="_blank" rel="noopener"
               data-ga="publication" data-title="Unsigned Matrix Completion (ISIT 2021)">
              Unsigned Matrix Completion
            </a>
          </div>
          <div class="pub-meta">
            <em>Yunzhen Yao</em>, Liangzu Peng, Manolis Tsakiris · ISIT 2021
          </div>
        </li>
      </ul>
    </section>

    <footer>
      © <span id="y"></span> Yunzhen Yao · Hosted on GitHub Pages
      <script>
        document.getElementById("y").textContent = new Date().getFullYear();
      </script>
    </footer>
  </div>

  <!-- GA4 click tracking (place before </body>) -->
  <script>
  document.addEventListener("DOMContentLoaded", function () {
    document.querySelectorAll("a[data-ga]").forEach(function (link) {
      link.addEventListener("click", function () {
        if (typeof gtag !== "function") return;

        const category = link.getAttribute("data-ga") || "";
        const title = link.getAttribute("data-title") || link.textContent.trim();
        const href = link.getAttribute("href") || "";

        gtag("event", "link_click", {
          event_category: category,
          event_label: title,
          link_url: href,
          outbound: /^https?:\/\//i.test(href) ? 1 : 0
        });
      });
    });
  });
  </script>
</body>
</html>