You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: _bibliography/references.bib
+4-2Lines changed: 4 additions & 2 deletions
Original file line number
Diff line number
Diff line change
@@ -30,7 +30,8 @@ @inproceedings{ pham2025the
30
30
url={https://openreview.net/forum?id=EEZLBhyer1},
31
31
pdf={https://openreview.net/pdf?id=EEZLBhyer1},
32
32
abstract={Sparse neural networks promise efficiency, yet training them effectively remains a fundamental challenge. Despite advances in pruning methods that create sparse architectures, understanding why some sparse structures are better trainable than others with the same level of sparsity remains poorly understood. Aiming to develop a systematic approach to this fundamental problem, we propose a novel theoretical framework based on the theory of graph limits, particularly graphons, that characterizes sparse neural networks in the infinite-width regime. Our key insight is that connectivity patterns of sparse neural networks induced by pruning methods converge to specific graphons as networks' width tends to infinity, which encodes implicit structural biases of different pruning methods. We postulate the Graphon Limit Hypothesis and provide empirical evidence to support it. Leveraging this graphon representation, we derive a Graphon Neural Tangent Kernel (Graphon NTK) to study the training dynamics of sparse networks in the infinite width limit. Graphon NTK provides a general framework for the theoretical analysis of sparse networks. We empirically show that the spectral analysis of Graphon NTK correlates with observed training dynamics of sparse networks, explaining the varying convergence behaviours of different pruning methods. Our framework provides theoretical insights into the impact of connectivity patterns on the trainability of various sparse network architectures.},
33
-
img={graphons.png}
33
+
img={graphons.png},
34
+
spotlight={true}
34
35
}
35
36
36
37
@inproceedings{jacobs2025mirror,
@@ -125,7 +126,8 @@ @inproceedings{
125
126
author={Shiwei Liu and Kai Han and Adriana Fernandez-Lopez and Ajay Kumar Jaiswal and Zahra Atashgahi and Boqian Wu and Edoardo Ponti and Callie Hao and Rebekka Burkholz and Olga Saukh and Lu Yin and Andreas Zinonos and Tianjin Huang and Jared Tanner and Yunhe Wang},
Copy file name to clipboardExpand all lines: _data/news.yml
+10-4Lines changed: 10 additions & 4 deletions
Original file line number
Diff line number
Diff line change
@@ -1,6 +1,12 @@
1
1
- date: 10. November 2025
2
2
headline: "Celia and Tom are presenting at the Workshop on Geometry, Topology, and Machine Learning ([GTML](https://www.mis.mpg.de/events/series/workshop-on-geometry-topology-and-machine-learning-gtml-2025)) in Leipzig."
3
3
4
+
- date: 29. October 2025
5
+
headline: "Celia is presenting her work on graph rewiring at the [GLOW](https://sites.google.com/view/graph-learning-on-weds/events) online seminar."
6
+
7
+
- date: 23. October 2025
8
+
headline: "Tom is presenting his work on implicit regularization at [AIM](https://aimath.nl/index.php/2025/08/27/phd-event/)'s event in Utrecht."
9
+
4
10
- date: 8. October 2025
5
11
headline: "Baraah is presenting her work on game-aware optimization at the [Shocklab](https://shocklab.net/seminars/) online seminar."
have been accepted at NeurIPS 2025, and [one](https://sites.google.com/wimlworkshop.org/wimlworkshopneurips2025/program) at the WiML Workshop."
16
22
17
23
- date: 15. September 2025
@@ -21,7 +27,7 @@
21
27
headline: "Tom is presenting his work on implicit regularization at [Cohere Labs](https://cohere.com/events/Cohere-Labs-Tom-Jacobs-2025): Open Science Community ([video](/outreach#tom-jacobs--cohere-labs-aug-14-2025))."
22
28
23
29
- date: 12. June 2025
24
-
headline: "Tom is attending the AI & Mathematics Workshop ([AIM](https://aimath.nl/index.php/2025/03/13/4th-aim-cluster-event-tilburg/)) at Tilburg University."
30
+
headline: "Tom is attending the AI & Mathematics Workshop ([AIM](https://aimath.nl/index.php/2025/03/13/4th-aim-cluster-event-tilburg/)) in Tilburg."
25
31
26
32
- date: 2. June 2025
27
33
headline: "Rebekka and Celia are presenting at [NetSci](https://netsci2025.github.io/) in Maastricht with a [satellite keynote](/outreach#netbiomed-2025-keynote-jun-2-2025) and a lightning talk."
@@ -39,7 +45,7 @@
39
45
headline: "Welcome Baraah!"
40
46
41
47
- date: 24. March 2025
42
-
headline: "Rebekka is at [CPAL](https://cpal.cc/spotlight_track/) presenting three [papers](/publications) as recent spotlights."
48
+
headline: "Rebekka is at [CPAL](https://cpal.cc/spotlight_track/) presenting three papers as recent spotlights."
43
49
44
50
- date: 13. February 2025
45
51
headline: "Celia is presenting her work on graph rewiring at [Cohere Labs](https://cohere.com/events/cohere-for-ai-Celia-Rubio-Madrigal-2025): Open Science Community ([video](/outreach#celia-rubio-madrigal--cohere-labs-feb-13-2025))."
@@ -54,7 +60,7 @@
54
60
headline: "Welcome to Gowtham and Nik!"
55
61
56
62
- date: 25. December 2024
57
-
headline: "Celia, Adarsh, Nimrah, and Rebekka are presenting their work at the [ELLIS Pre-NeurIPS](https://lacoco-lab.github.io/events/PreNeurIPS2024/) Session in Saarbrücken."
63
+
headline: "Celia, Adarsh, Nimrah, and Rebekka have posters at the [ELLIS Pre-NeurIPS](https://lacoco-lab.github.io/events/PreNeurIPS2024/) Session in Saarbrücken."
Copy file name to clipboardExpand all lines: _site/index.html
+8-8Lines changed: 8 additions & 8 deletions
Original file line number
Diff line number
Diff line change
@@ -135,23 +135,23 @@ <h3>News</h3>
135
135
136
136
<hr/>
137
137
138
-
<b>08 Oct 2025</b>
139
-
<p>Baraah is presenting her work on game-aware optimization at the <ahref="https://shocklab.net/seminars/">Shocklab</a> online seminar.</p>
138
+
<b>29 Oct 2025</b>
139
+
<p>Celia is presenting her work on graph rewiring at the <ahref="https://sites.google.com/view/graph-learning-on-weds/events">GLOW</a> online seminar.</p>
140
140
141
141
<hr/>
142
142
143
-
<b>06 Oct 2025</b>
144
-
<p>Welcome to Adnan (University of Calgary) and Levy (University of Campinas) for research stays. Furthermore, Prof. <ahref="https://yani.ai/">Yani Ioannou</a> (University of Calgary) is visiting us for a talk on sparse training.</p>
143
+
<b>23 Oct 2025</b>
144
+
<p>Tom is presenting his work on implicit regularization at <ahref="https://aimath.nl/index.php/2025/08/27/phd-event/">AIM</a>’s event in Utrecht.</p>
145
145
146
146
<hr/>
147
147
148
-
<b>18 Sep 2025</b>
149
-
<p>Three papers <ahref="https://openreview.net/forum?id=XKnOA7MhCz">(1)</a><ahref="https://openreview.net/forum?id=iwKT7MEZZw">(2)</a><ahref="https://openreview.net/forum?id=EEZLBhyer1">(3)</a> have been accepted at NeurIPS 2025, and <ahref="https://sites.google.com/wimlworkshop.org/wimlworkshopneurips2025/program">one</a>at the WiML Workshop.</p>
148
+
<b>08 Oct 2025</b>
149
+
<p>Baraah is presenting her work on game-aware optimization at the <ahref="https://shocklab.net/seminars/">Shocklab</a>online seminar.</p>
150
150
151
151
<hr/>
152
152
153
-
<b>15 Sep 2025</b>
154
-
<p>Rebekka and Celia are presenting at the Workshop on Mining and Learning with Graphs (<ahref="https://mlg-europe.github.io/2025/">MLG</a>) in Porto with a keynote and two posters, respectively.</p>
153
+
<b>06 Oct 2025</b>
154
+
<p>Welcome to Adnan (University of Calgary) and Levy (University of Campinas) for research stays. Furthermore, Prof. <ahref="https://yani.ai/">Yani Ioannou</a> (University of Calgary) is visiting us for a talk on sparse training.</p>
Copy file name to clipboardExpand all lines: _site/news.html
+14-4Lines changed: 14 additions & 4 deletions
Original file line number
Diff line number
Diff line change
@@ -82,6 +82,16 @@ <h4>10 November 2025</h4>
82
82
83
83
<hr/>
84
84
85
+
<h4>29 October 2025</h4>
86
+
<p>Celia is presenting her work on graph rewiring at the <ahref="https://sites.google.com/view/graph-learning-on-weds/events">GLOW</a> online seminar.</p>
87
+
88
+
<hr/>
89
+
90
+
<h4>23 October 2025</h4>
91
+
<p>Tom is presenting his work on implicit regularization at <ahref="https://aimath.nl/index.php/2025/08/27/phd-event/">AIM</a>’s event in Utrecht.</p>
92
+
93
+
<hr/>
94
+
85
95
<h4>08 October 2025</h4>
86
96
<p>Baraah is presenting her work on game-aware optimization at the <ahref="https://shocklab.net/seminars/">Shocklab</a> online seminar.</p>
87
97
@@ -93,7 +103,7 @@ <h4>06 October 2025</h4>
93
103
<hr/>
94
104
95
105
<h4>18 September 2025</h4>
96
-
<p>Three papers <ahref="https://openreview.net/forum?id=XKnOA7MhCz">(1)</a><ahref="https://openreview.net/forum?id=iwKT7MEZZw">(2)</a><ahref="https://openreview.net/forum?id=EEZLBhyer1">(3)</a> have been accepted at NeurIPS 2025, and <ahref="https://sites.google.com/wimlworkshop.org/wimlworkshopneurips2025/program">one</a> at the WiML Workshop.</p>
106
+
<p>Three papers <ahref="https://openreview.net/forum?id=XKnOA7MhCz">(1)</a><ahref="https://openreview.net/forum?id=iwKT7MEZZw">(2)</a><ahref="https://openreview.net/forum?id=EEZLBhyer1">(3-Spotlight)</a> have been accepted at NeurIPS 2025, and <ahref="https://sites.google.com/wimlworkshop.org/wimlworkshopneurips2025/program">one</a> at the WiML Workshop.</p>
97
107
98
108
<hr/>
99
109
@@ -108,7 +118,7 @@ <h4>14 August 2025</h4>
108
118
<hr/>
109
119
110
120
<h4>12 June 2025</h4>
111
-
<p>Tom is attending the AI & Mathematics Workshop (<ahref="https://aimath.nl/index.php/2025/03/13/4th-aim-cluster-event-tilburg/">AIM</a>) at Tilburg University.</p>
121
+
<p>Tom is attending the AI & Mathematics Workshop (<ahref="https://aimath.nl/index.php/2025/03/13/4th-aim-cluster-event-tilburg/">AIM</a>) in Tilburg.</p>
112
122
113
123
<hr/>
114
124
@@ -138,7 +148,7 @@ <h4>01 May 2025</h4>
138
148
<hr/>
139
149
140
150
<h4>24 March 2025</h4>
141
-
<p>Rebekka is at <ahref="https://cpal.cc/spotlight_track/">CPAL</a> presenting three <ahref="/publications">papers</a> as recent spotlights.</p>
151
+
<p>Rebekka is at <ahref="https://cpal.cc/spotlight_track/">CPAL</a> presenting three papers as recent spotlights.</p>
142
152
143
153
<hr/>
144
154
@@ -158,7 +168,7 @@ <h4>01 December 2024</h4>
158
168
<hr/>
159
169
160
170
<h4>25 December 2024</h4>
161
-
<p>Celia, Adarsh, Nimrah, and Rebekka are presenting their work at the <ahref="https://lacoco-lab.github.io/events/PreNeurIPS2024/">ELLIS Pre-NeurIPS</a> Session in Saarbrücken.</p>
171
+
<p>Celia, Adarsh, Nimrah, and Rebekka have posters at the <ahref="https://lacoco-lab.github.io/events/PreNeurIPS2024/">ELLIS Pre-NeurIPS</a> Session in Saarbrücken.</p>
<spanclass="na">author</span><spanclass="p">=</span><spanclass="s">{Liu, Shiwei and Han, Kai and Fernandez-Lopez, Adriana and Jaiswal, Ajay Kumar and Atashgahi, Zahra and Wu, Boqian and Ponti, Edoardo and Hao, Callie and Burkholz, Rebekka and Saukh, Olga and Yin, Lu and Zinonos, Andreas and Huang, Tianjin and Tanner, Jared and Wang, Yunhe}</span><spanclass="p">,</span>
0 commit comments