372 citations · 641 across the 71 of their papers we have counts for
95 papers
Spectral Geometry of Attention: From Information Routing to Uncertainty
Giulio Viganò, Simone Melzi, Maks Ovsjanikov
In this work, we study transformer attention through the lens of spectral geometry and operator theory. We view each attention head as a functional map between Hilbert spaces of fu…
Pattern Formation in Transformers
Erkan Turan, Gaspard Abel, Maks Ovsjanikov
What are the inductive biases of a Transformer architecture? Existing theory on how the forward pass shapes representations either considers whether Transformers escape from rank c…
3DHarnessBench: Probing Agentic 3D-to-Code Capabilities of Frontier Vision-Language Models
Ling Liu, Bingchen Gong, Amal Dev Parakkat +1
We introduce 3DHarnessBench, a benchmark that evaluates the agentic ability of frontier vision-language models (VLMs) to recover 3D geometry as Blender Python code from a variety o…
The Computational Basis of Confidence in Large Language Models
Dharshan Kumaran, Viorica Patraucean, Maks Ovsjanikov +2
Reliable confidence -- the probability that a model's own answer is correct -- is essential for the trustworthy deployment of language models. Existing work has largely evaluated c…
SeeSE3: Emergence of 3D Space in Vision Features
Caroline Chen, Sayna Ebrahimi, Fedor Kitashov +4
In this paper, we ask whether vision foundation models construct representations that reflect the intrinsic properties of 3D Euclidean space. Unlike previous works that probe 3D aw…
Gen4U: Unifying Video Generation and Understanding via Diffusion
Michael King, Aravindh Mahendran, Matthew Koichi Grimes +5
Prior work suggests that diffusion representations capture low-level geometry but struggle with high-level semantics. We demonstrate that state-of-the-art video diffusion models ov…
The Art of Interrogation: Consistency Amplifies Factuality in Spatial Reasoning
Theo Uscidda, Marta Tintore Gazulla, Maks Ovsjanikov +2
Current Large Reasoning Models (LRMs) exhibit remarkable general capabilities but significantly underperform in spatial reasoning tasks. Existing approaches treat this gap as a kno…
FILTR: Extracting Topological Features from Pretrained 3D Models
Louis Martinez, Maks Ovsjanikov
Recent advances in pretraining 3D point cloud encoders (e.g., Point-BERT, Point-MAE) have produced powerful models, whose abilities are typically evaluated on geometric or semantic…
From Blobs to Spokes: High-Fidelity Surface Reconstruction via Oriented Gaussians
Diego Gomez, Antoine Guédon, Nissim Maruani +2
3D Gaussian Splatting (3DGS) has revolutionized fast novel view synthesis, yet its opacity-based formulation makes surface extraction fundamentally difficult. Unlike implicit metho…
Beyond Prompts: Unconditional 3D Inversion for Out-of-Distribution Shapes
Victoria Yue Chen, Emery Pierson, Léopold Maillard +1
Text-driven inversion of generative models is a core paradigm for manipulating 2D or 3D content, unlocking numerous applications such as text-based editing, style transfer, or inve…
SceneTeract: Probing and Improving Agent-Aware Activity Reasoning in 3D Indoor Scenes
Léopold Maillard, Francis Engelmann, Tom Durand +5
Indoor 3D scenes are ultimately meant to be used: an embodied agent should be able to navigate, reach objects, and complete diverse activities. Yet whether a given scene actually s…
Generative Drifting is Secretly Score Matching: a Spectral and Variational Perspective
Erkan Turan, Nicolas Dufour, Maks Ovsjanikov
Generative Modeling via Drifting~\citep{deng2026drifting} has recently achieved state-of-the-art one-step image generation through a kernel-based drift operator, yet its success is…
A Mixed Diet Makes DINO An Omnivorous Vision Encoder
Rishabh Kabra, Maks Ovsjanikov, Drew A. Hudson +5
Pre-trained vision encoders like DINOv2 have demonstrated exceptional performance on unimodal tasks. However, we observe that their features are poorly aligned across different vis…
Beyond ReLU: Bifurcation, Oversmoothing, and Topological Priors
Erkan Turan, Gaspard Abel, Maysam Behmanesh +2
Graph Neural Networks (GNNs) learn node representations through iterative network-based message-passing. While powerful, deep GNNs suffer from oversmoothing, where node features co…
PatchAlign3D: Local Feature Alignment for Dense 3D Shape Understanding
Souhail Hadgi, Bingchen Gong, Ramana Sundararaman +4
Current foundation models for 3D shapes excel at global tasks (retrieval, classification) but transfer poorly to local part-level reasoning. Recent approaches leverage vision and l…
Unified Brain Surface and Volume Registration
S. Mazdak Abulnaga, Andrew Hoopes, Malte Hoffmann +6
Accurate registration of brain MRI scans is fundamental for cross-subject analysis in neuroscientific studies. This involves aligning both the cortical surface of the brain and the…
Unique Lives, Shared World: Learning from Single-Life Videos
Tengda Han, Sayna Ebrahimi, Dilara Gokay +8
We introduce the "single-life" learning paradigm, where we train a distinct vision model exclusively on egocentric videos captured by one individual. We leverage the multiple viewp…
Dynamic Reflections: Probing Video Representations with Text Alignment
Tyler Zhu, Tengda Han, Leonidas Guibas +2
The alignment of representations from different modalities has recently been shown to provide insights on the structural similarities and downstream capabilities of different encod…
Graph Alignment via Dual-Pass Spectral Encoding and Latent Space Communication
Maysam Behmanesh, Erkan Turan, Maks Ovsjanikov
Graph alignment, the problem of identifying corresponding nodes across multiple graphs, is fundamental to numerous applications. Most existing unsupervised methods embed node featu…
LACONIC: A 3D Layout Adapter for Controllable Image Creation
Léopold Maillard, Tom Durand, Adrien Ramanana Rahary +1
Existing generative approaches for guided image synthesis of multi-object scenes typically rely on 2D controls in the image or text space. As a result, these methods struggle to ma…
DiffuMatch: Category-Agnostic Spectral Diffusion Priors for Robust Non-rigid Shape Matching
Emery Pierson, Lei Li, Angela Dai +1
Deep functional maps have recently emerged as a powerful tool for solving non-rigid shape correspondence tasks. Methods that use this approach combine the power and flexibility of…
SciVid: Cross-Domain Evaluation of Video Models in Scientific Applications
Yana Hasson, Pauline Luc, Liliane Momeni +10
In recent years, there has been a proliferation of spatiotemporal foundation models in different scientific disciplines. While promising, these models are often domain-specific and…
Frozen Forecasting: A Unified Evaluation
Jacob C Walker, Pedro Vélez, Luisa Polania Cabrera +7
Forecasting future events is a fundamental capability for general-purpose systems that plan or act across different levels of abstraction. Yet, evaluating whether a forecast is "co…
MILo: Mesh-In-the-Loop Gaussian Splatting for Detailed and Efficient Surface Reconstruction
Antoine Guédon, Diego Gomez, Nissim Maruani +3
While recent advances in Gaussian Splatting have enabled fast reconstruction of high-quality 3D scenes from images, extracting accurate surface meshes remains a challenge. Current…