REVIEW 12 cited by
Frame Averaging for Invariant and Equivariant Network Design
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Frame Averaging for Invariant and Equivariant Network Design
read the original abstract
Many machine learning tasks involve learning functions that are known to be invariant or equivariant to certain symmetries of the input data. However, it is often challenging to design neural network architectures that respect these symmetries while being expressive and computationally efficient. For example, Euclidean motion invariant/equivariant graph or point cloud neural networks. We introduce Frame Averaging (FA), a general purpose and systematic framework for adapting known (backbone) architectures to become invariant or equivariant to new symmetry types. Our framework builds on the well known group averaging operator that guarantees invariance or equivariance but is intractable. In contrast, we observe that for many important classes of symmetries, this operator can be replaced with an averaging operator over a small subset of the group elements, called a frame. We show that averaging over a frame guarantees exact invariance or equivariance while often being much simpler to compute than averaging over the entire group. Furthermore, we prove that FA-based models have maximal expressive power in a broad setting and in general preserve the expressive power of their backbone architectures. Using frame averaging, we propose a new class of universal Graph Neural Networks (GNNs), universal Euclidean motion invariant point cloud networks, and Euclidean motion invariant Message Passing (MP) GNNs. We demonstrate the practical effectiveness of FA on several applications including point cloud normal estimation, beyond $2$-WL graph separation, and $n$-body dynamics prediction, achieving state-of-the-art results in all of these benchmarks.
Forward citations
Cited by 12 Pith papers
-
EquiVLA: A General Framework for Rotationally Equivariant Vision-Language-Action Models
EquiVLA is the first general framework for end-to-end SO(2)-equivariant VLA models using EquiPerceptor and EquiActor modules, reporting improved success rates on LIBERO, CALVIN, and real-robot benchmarks.
-
Aligning Network Equivariance with Data Symmetry: A Theoretical Framework and Adaptive Approach for Image Restoration
A new dataset-level non-strict symmetry measure allows deriving bounded equivariance for restoration models and motivates an adaptive network that aligns with per-sample symmetry to reduce expected risk.
-
When and How to Canonize: A Generalization Perspective
Canonization produces generalization bounds ranging from invariant-optimal to non-invariant depending on regularity, with Hilbert-curve ordering proven to give polynomial covering-number growth for point clouds while ...
-
Parametrizing Convex Sets Using Sublinear Neural Networks
Sublinear neural networks parametrize convex sets by learning their support and gauge functions, backed by a universal approximation theorem and tested on shape optimization tasks.
-
Inverting Data Transformations via Diffusion Sampling
A Lie-group diffusion sampler that inverts unknown data transformations at test time, using only an energy function, and improves pretrained models on affine/homography images and PDE solving.
-
Group-Equivariant Poincar\'e Convolutional Networks
Equivariant Poincaré ResNets combine hyperbolic geometry with C4 and D4 group symmetries via specialized reshaping, permutations, and batch norm to reduce optimization space and speed convergence while staying inside ...
-
Platonic Transformers: A Solid Choice For Equivariance
Platonic Transformers achieve exact equivariance to translations plus discrete Platonic-solid rotations by lifting features into multiple reference frames and sharing one RoPE attention across them, with a linear-time...
-
Adaptive Canonicalization with Application to Invariant Anisotropic Geometric Networks
Adaptive canonicalization selects input canonical forms by maximizing network predictive confidence to yield continuous symmetry-preserving models with universal approximation for equivariant geometric networks.
-
Graph Mamba Operator: A Latent Simulator for Interacting Particle Systems
GraMO couples graph interactions and temporal state updates in one linear recurrence with input-dependent coefficients to simulate N-body, motion, and robotics systems with lower long-horizon error than prior GNN or S...
-
Multi-Alignment Contrastive Learning for Enzyme--Reaction Retrieval
FGW-CLIP, a contrastive method that aligns enzymes and reactions while also aligning within-domain EC structure with a Gromov-Wasserstein regularizer, reports state-of-the-art retrieval on EnzymeMap and ReactZyme.
-
Explaining the Explainers in Graph Neural Networks: a Comparative Study
Benchmark study of ten GNN explainers on eight architectures and six datasets that isolates usable components and issues practical recommendations.
-
TAGTorch: A PyTorch Library for Geometry, Topology, and Symmetry-Aware Machine Learning
TAGTorch packages geometry, topology, and symmetry-aware deep learning tools into one PyTorch library, with a design built around unified data transforms and group representations.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.