Function graph transformers use graph measures to provide a measure-theoretic framework where standard transformer components universally approximate operators between function spaces while preserving single-valued function outputs.
Walrus: A cross-domain foundation model for continuum dynamics
11 Pith papers cite this work. Polarity classification is still indexing.
years
2026 11verdicts
UNVERDICTED 11representative citing papers
Breakeven complexity is introduced to evaluate neural PDE solvers by total end-to-end cost, with results indicating they become advantageous for harder problems such as higher dimensions, longer rollouts, and higher Reynolds numbers.
Fine-tuning neural PDE operators to regime endpoints reveals a physical direction in weight space that CCM uses to compose accurate merged models for new or extrapolated regimes from metadata or short prefixes.
OmniMol transfers a billion-jet pre-trained PET foundation model from HEP to molecular dynamics via an interaction-matrix attention bias, delivering strong performance on the oMol dataset with minimal fine-tuning and fast inference.
WaveLiT combines wavelet tokenization, linear attention, and multiscale pyramids to produce parameter-efficient neural PDE solvers that match much larger models on TheWell benchmarks.
AOT-POT adaptively reshapes complex PDE solution operators via input-dependent transformations and parallel stream mixing to enable effective large-scale pre-training, yielding SOTA results on 12 benchmarks with minimal added parameters.
Small transformers learn to forecast unseen dynamical systems in-context by using delay embeddings to recover the manifold and forecasting its invariant sets via a transfer-operator strategy.
Case study applies SAE probing with enstrophy triage to a continuum-dynamics foundation model and reports intermittent feature consistency that does not align with standard physics while linking some output discrepancies to specific feature changes.
jNO introduces a unified JAX tracing system for data-driven and physics-informed neural operator training that compiles domains, residuals, losses, and diagnostics into one pipeline.
A replay-based continual learning strategy for physics-informed neural operators mitigates catastrophic forgetting on prior physical problems while enabling efficient adaptation to new data using only physical constraints.
The paper reviews data sources, physical models, downstream applications, and AI techniques to outline considerations for building a foundation model for the Martian atmosphere.
citing papers explorer
-
Function graph transformers universally approximate operators between function spaces
Function graph transformers use graph measures to provide a measure-theoretic framework where standard transformer components universally approximate operators between function spaces while preserving single-valued function outputs.
-
Breakeven complexity: A new perspective on neural partial differential equation solvers
Breakeven complexity is introduced to evaluate neural PDE solvers by total end-to-end cost, with results indicating they become advantageous for harder problems such as higher dimensions, longer rollouts, and higher Reynolds numbers.
-
Discovering Physical Directions in Weight Space: Composing Neural PDE Experts
Fine-tuning neural PDE operators to regime endpoints reveals a physical direction in weight space that CCM uses to compose accurate merged models for new or extrapolated regimes from metadata or short prefixes.
-
OmniMol: Transferring Particle Physics Knowledge to Molecular Dynamics with Point-Edge Transformers
OmniMol transfers a billion-jet pre-trained PET foundation model from HEP to molecular dynamics via an interaction-matrix attention bias, delivering strong performance on the oMol dataset with minimal fine-tuning and fast inference.
-
Small Models, Strong Priors: Architectural Inductive Bias for Parameter-Efficient Neural PDE Solvers
WaveLiT combines wavelet tokenization, linear attention, and multiscale pyramids to produce parameter-efficient neural PDE solvers that match much larger models on TheWell benchmarks.
-
AOT-POT: Adaptive Operator Transformation for Large-Scale PDE Pre-training
AOT-POT adaptively reshapes complex PDE solution operators via input-dependent transformations and parallel stream mixing to enable effective large-scale pre-training, yielding SOTA results on 12 benchmarks with minimal added parameters.
-
Transformers for dynamical systems learn transfer operators in-context
Small transformers learn to forecast unseen dynamical systems in-context by using delay embeddings to recover the manifold and forecasting its invariant sets via a transfer-operator strategy.
-
Sparse probes and murky physics: a case study of interpretability challenges in a foundation model for continuum dynamics
Case study applies SAE probing with enstrophy triage to a continuum-dynamics foundation model and reports intermittent feature consistency that does not align with standard physics while linking some output discrepancies to specific feature changes.
-
jNO: A JAX Library for Neural Operator and Foundation Model Training
jNO introduces a unified JAX tracing system for data-driven and physics-informed neural operator training that compiles domains, residuals, losses, and diagnostics into one pipeline.
-
Replay-Based Continual Learning for Physics-Informed Neural Operators
A replay-based continual learning strategy for physics-informed neural operators mitigates catastrophic forgetting on prior physical problems while enabling efficient adaptation to new data using only physical constraints.
-
Towards a Foundation Model for the Martian Atmosphere
The paper reviews data sources, physical models, downstream applications, and AI techniques to outline considerations for building a foundation model for the Martian atmosphere.