Skip to content

Commit 398bb9d

Browse files
committed
cvpr
1 parent fe0a15b commit 398bb9d

14 files changed

Lines changed: 296 additions & 21 deletions

File tree

_bibliography/references.bib

Lines changed: 46 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,25 @@
11
---
22
---
33
4+
@inproceedings{merge,
5+
title={Bridging Domains through Subspace-Aware Model Merging},
6+
author={Levy Chaves and Chao Zhou and Rebekka Burkholz and Eduardo Valle and Andra Avila},
7+
year={2026},
8+
booktitle={IEEE Conference on Computer Vision and Pattern Recognition},
9+
img={model-merging.png},
10+
}
11+
12+
@inproceedings{sanyal2026games,
13+
title={Frequency-Based Hyperparameter Selection in Games},
14+
author={Aniket Sanyal and Baraah A. M. Sidahmed and Rebekka Burkholz and Tatjana Chavdarova},
15+
year={2026},
16+
booktitle={International Conference on Artificial Intelligence and Statistics},
17+
img={frequency-based-hyperparameter-selection.png},
18+
url={https://arxiv.org/abs/2601.18409},
19+
pdf={https://arxiv.org/pdf/2601.18409},
20+
abstract={Learning in smooth games fundamentally differs from standard minimization due to rotational dynamics, which invalidate classical hyperparameter tuning strategies. Despite their practical importance, effective methods for tuning in games remain underexplored. A notable example is LookAhead (LA), which achieves strong empirical performance but introduces additional parameters that critically influence performance. We propose a principled approach to hyperparameter selection in games by leveraging frequency estimation of oscillatory dynamics. Specifically, we analyze oscillations both in continuous-time trajectories and through the spectrum of the discrete dynamics in the associated frequency-based space. Building on this analysis, we introduce \emph{Modal LookAhead (MoLA)}, an extension of LA that selects the hyperparameters adaptively to a given problem. We provide convergence guarantees and demonstrate in experiments that MoLA accelerates training in both purely rotational games and mixed regimes, all with minimal computational overhead.}
21+
}
22+
423
@inproceedings{
524
reddy2026boosting,
625
title={Boosting for Predictive Sufficiency},
@@ -55,10 +74,23 @@ @inproceedings{
5574
author={Celia Rubio-Madrigal and Rebekka Burkholz},
5675
booktitle={Women in Machine Learning Workshop @ NeurIPS},
5776
year={2025},
58-
url={https://openreview.net/forum?id=OHgWEMce80},
59-
pdf={https://openreview.net/pdf?id=OHgWEMce80},
77+
url={https://arxiv.org/abs/2601.19449},
78+
pdf={https://arxiv.org/pdf/2601.19449},
6079
img={fixed-aggregation-features.png},
6180
abstract={Graph neural networks (GNNs) are widely believed to excel at node representation learning through trainable neighborhood aggregations. We challenge this view by introducing Fixed Aggregation Features (FAFs), a training-free approach that transforms graph learning tasks into tabular problems. This simple shift enables the use of well-established tabular methods, offering strong interpretability and the flexibility to deploy diverse classifiers. Across 14 benchmarks, well-tuned multilayer perceptrons trained on FAFs rival or outperform state-of-the-art GNNs and graph transformers on 12 tasks -- often using only mean aggregation. The only exceptions are the Roman Empire and Minesweeper datasets, which typically require unusually deep GNNs. To explain the theoretical possibility of non-trainable aggregations, we connect our findings to Kolmogorov–Arnold representations and discuss when mean aggregation can be sufficient. In conclusion, our results call for (i) richer benchmarks benefiting from learning diverse neighborhood aggregations, (ii) strong tabular baselines as standard, and (iii) employing and advancing tabular models for graph data to gain new insights into related tasks.},
81+
code={https://github.com/celrm/fixed-aggregation-features}
82+
}
83+
84+
@inproceedings{
85+
gadhikar2025optrot,
86+
title={OptRot: Mitigating Weight Outliers via Data-Free Rotations for Post-Training Quantization},
87+
author={Advait Gadhikar and Riccardo Grazzi and James Hensman},
88+
booktitle={Machine Learning for Systems @ NeurIPS},
89+
year={2025},
90+
img={optrot.png},
91+
url={https://openreview.net/forum?id=4uwRaBhjHY},
92+
pdf={https://openreview.net/pdf?id=4uwRaBhjHY},
93+
abstract={We introduce OptRot, a data-free preprocessing method to learn fusible rotations for post-training quantization of language models. OptRot reduces weight outliers by finding rotations which minimize the element-wise fourth power of the rotated weights. We show how reducing weight outliers can provably improve weight quantization performance and how OptRot rotations can outperform both Hadamard rotations and rotations learned by the data-dependent method SpinQuant.},
6294
}
6395

6496
@inproceedings{ zhou2025payattentionsmallweights,
@@ -106,6 +138,18 @@ @inproceedings{jacobs2025mirror
106138
abstract={Implicit bias plays an important role in explaining how overparameterized models generalize well. Explicit regularization like weight decay is often employed in addition to prevent overfitting. While both concepts have been studied separately, in practice, they often act in tandem. Understanding their interplay is key to controlling the shape and strength of implicit bias, as it can be modified by explicit regularization. To this end, we incorporate explicit regularization into the mirror flow framework and analyze its lasting effects on the geometry of the training dynamics, covering three distinct effects: positional bias, type of bias, and range shrinking. Our analytical approach encompasses a broad class of problems, including sparse coding, matrix sensing, single-layer attention, and LoRA, for which we demonstrate the utility of our insights. To exploit the lasting effect of regularization and highlight the potential benefit of dynamic weight decay schedules, we propose to switch off weight decay during training, which can improve generalization, as we demonstrate in experiments.},
107139
}
108140

141+
@inproceedings{
142+
gadhikar2025attention,
143+
title={Attention Is All You Need For Mixture-of-Depths Routing},
144+
author={Advait Gadhikar and Souptik Kumar Majumdar and Niclas Popp and Piyapat Saranrittichai and Martin Rapp and Lukas Schott},
145+
booktitle={Workshop on Scalable Optimization for Efficient and Adaptive Foundation Models @ ICLR},
146+
year={2025},
147+
img={mod-routing.png},
148+
url={https://openreview.net/forum?id=1uDP4ld3eZ},
149+
pdf={https://openreview.net/pdf?id=1uDP4ld3eZ},
150+
abstract={Advancements in deep learning are driven by training models with increasingly larger numbers of parameters, which in turn heightens the computational demands. To address this issue, Mixture-of-Depths (MoD) models have been proposed to dynamically focus computations on the most relevant parts of the inputs, thereby enabling the deployment of large-parameter models with high efficiency during inference and training. However, conventional MoD models employ additional network layers specifically for the routing which are difficult to train, and add complexity to the model. In this paper, we introduce a novel attention-based routing mechanism A-MoD that leverages the existing attention map of the preceding layer for routing decisions within the current layer. Compared to standard routing, A-MoD allows for more efficient training as it introduces no additional trainable parameters and can be easily adapted from pre-trained transformer models. Furthermore, it can increase the performance of the MoD model. For instance, we observe up to 2% higher accuracy on ImageNet compared to standard routing and isoFLOP ViT baselines.}
151+
}
152+
109153
@inproceedings{
110154
jacobs2025mask,
111155
title={Mask in the Mirror: Implicit Sparsification},

_data/news.yml

Lines changed: 10 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -5,14 +5,20 @@
55
[(3)](https://openreview.net/forum?id=sFjxg8cyJS)
66
[(4)](https://openreview.net/forum?id=1mQT8PXIy8)."
77

8+
- date: 31. March 2026
9+
headline: "Congratulations to Advait for submitting his PhD!"
10+
811
- date: 23. March 2026
9-
headline: "💬 Chao, Celia, and Tom are presenting at [CPAL](https://cpal.cc/spotlight_track/) with five papers as recent spotlights."
12+
headline: "💬 Most of the group is presenting at [CPAL](https://cpal.cc/spotlight_track/) with six papers as recent spotlights."
1013

1114
- date: 1. March 2026
12-
headline: "Congratulations to Advait for finishing his PhD!"
15+
headline: "Welcome to Rohan, Franka, and Jonas!"
16+
17+
- date: 20. February 2026
18+
headline: "📛 We have a paper on model merging accepted at CVPR 2026, and a [paper](https://arxiv.org/pdf/2601.18409) on learning on games accepted at AISTATS 2026."
1319

14-
- date: 1. February 2026
15-
headline: "Welcome to Rohan and Franka!"
20+
- date: 9. February 2026
21+
headline: "💬 Tom is presenting at the [Tropical mathematics and ML](https://researchseminars.org/talk/TropicalmathandML/32/) online seminar."
1622

1723
- date: 3. December 2025
1824
headline: "📛 Three papers have been accepted at NeurIPS 2025

_layouts/bib.html

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -128,6 +128,10 @@
128128
<a class="conf btn btn-sm z-depth-0">ICML</a>
129129
{%- elsif entrytype contains "International Conference on Learning Representations" -%}
130130
<a class="conf btn btn-sm z-depth-0">ICLR</a>
131+
{%- elsif entrytype contains "International Conference on Artificial Intelligence and Statistics" -%}
132+
<a class="conf btn btn-sm z-depth-0">AISTATS</a>
133+
{%- elsif entrytype contains "Conference on Computer Vision and Pattern Recognition" -%}
134+
<a class="conf btn btn-sm z-depth-0">CVPR</a>
131135
{%- endif -%}
132136

133137
{%- if entry.spotlight -%}
271 KB
Loading
37.7 KB
Loading
110 KB
Loading
80.9 KB
Loading

_site/index.html

Lines changed: 8 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -133,23 +133,23 @@ <h3 style="display:inline-block; margin:0em 0.5em 0.5em 0em; vertical-align:bott
133133

134134
<hr/>
135135

136-
<b>23 Mar 2026</b>
137-
<p>💬 Chao, Celia, and Tom are presenting at <a href="https://cpal.cc/spotlight_track/">CPAL</a> with five papers as recent spotlights.</p>
136+
<b>31 Mar 2026</b>
137+
<p>Congratulations to Advait for submitting his PhD!</p>
138138

139139
<hr/>
140140

141-
<b>01 Mar 2026</b>
142-
<p>Congratulations to Advait for finishing his PhD!</p>
141+
<b>23 Mar 2026</b>
142+
<p>💬 Most of the group is presenting at <a href="https://cpal.cc/spotlight_track/">CPAL</a> with six papers as recent spotlights.</p>
143143

144144
<hr/>
145145

146-
<b>01 Feb 2026</b>
147-
<p>Welcome to Rohan and Franka!</p>
146+
<b>01 Mar 2026</b>
147+
<p>Welcome to Rohan, Franka, and Jonas!</p>
148148

149149
<hr/>
150150

151-
<b>03 Dec 2025</b>
152-
<p>📛 Three papers have been accepted at NeurIPS 2025 <a href="https://openreview.net/forum?id=XKnOA7MhCz">(1)</a> <a href="https://openreview.net/forum?id=iwKT7MEZZw">(2)</a> <a href="https://openreview.net/forum?id=EEZLBhyer1">(3-Spotlight)</a>, and four more at workshops <a href="https://openreview.net/forum?id=OHgWEMce80">(1)</a> <a href="https://sites.google.com/view/dynafrontneurips25/papers">(2)</a> <a href="https://openreview.net/forum?id=i0rFGQBGzs">(3)</a> <a href="https://openreview.net/forum?id=4uwRaBhjHY">(4)</a>.</p>
151+
<b>20 Feb 2026</b>
152+
<p>📛 We have a paper on model merging accepted at CVPR 2026, and a <a href="https://arxiv.org/pdf/2601.18409">paper</a> on learning on games accepted at AISTATS 2026.</p>
153153

154154

155155

_site/news.html

Lines changed: 14 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -83,18 +83,28 @@ <h4>23 April 2026</h4>
8383

8484
<hr />
8585

86+
<h4>31 March 2026</h4>
87+
<p>Congratulations to Advait for submitting his PhD!</p>
88+
89+
<hr />
90+
8691
<h4>23 March 2026</h4>
87-
<p>💬 Chao, Celia, and Tom are presenting at <a href="https://cpal.cc/spotlight_track/">CPAL</a> with five papers as recent spotlights.</p>
92+
<p>💬 Most of the group is presenting at <a href="https://cpal.cc/spotlight_track/">CPAL</a> with six papers as recent spotlights.</p>
8893

8994
<hr />
9095

9196
<h4>01 March 2026</h4>
92-
<p>Congratulations to Advait for finishing his PhD!</p>
97+
<p>Welcome to Rohan, Franka, and Jonas!</p>
98+
99+
<hr />
100+
101+
<h4>20 February 2026</h4>
102+
<p>📛 We have a paper on model merging accepted at CVPR 2026, and a <a href="https://arxiv.org/pdf/2601.18409">paper</a> on learning on games accepted at AISTATS 2026.</p>
93103

94104
<hr />
95105

96-
<h4>01 February 2026</h4>
97-
<p>Welcome to Rohan and Franka!</p>
106+
<h4>09 February 2026</h4>
107+
<p>💬 Tom is presenting at the <a href="https://researchseminars.org/talk/TropicalmathandML/32/">Tropical mathematics and ML</a> online seminar.</p>
98108

99109
<hr />
100110

0 commit comments

Comments
 (0)