ffortissimo is a JAX-based freeform forward-modeling pipeline that fits complex dust distributions and infers scattering properties in KLIP-reduced images of circumstellar disks such as HR 4796A.
hub Mixed citations
On the limited memory BFGS method for large scale optimization
Mixed citation behavior. Most common role is background (57%).
27
Pith papers citing it
8,529
external citations · OpenAlex
Background
57% of classified citations
hub tools
citation-role summary
background 4
method 3
citation-polarity summary
representative citing papers
The curvature-aware precision controller adapts between FP32 and FP64 during PINN training to match double-precision accuracy at reduced computational cost.
SGD, approximations of Newton's method, natural gradient descent, and Adam are proven compatible with evolutionary dynamics when augmented with DLS noise, turning them into valid in silico simulations of asexual Darwinian evolution.
An R package that unifies spatio-temporal STARMA and GARCH-type models through double generalized linear models, allowing simultaneous modeling of means and dispersion for count or continuous spatial time series.
PI-MFA optimizes tensor-product B-spline control points to balance data fidelity against PDE residuals, producing physically consistent continuous flow fields.
Reasoning Arena converts non-diverse reward groups in RLVR into relative rewards via adaptive trace tournaments and Bradley-Terry fitting on anchor comparisons, claiming 7.6% average gains and 27-41% faster training on math/coding benchmarks.
A proximal limited-memory quasi-Newton scheme is developed for nonsmooth nonconvex optimization, with global convergence proven under mild assumptions and rates under the Kurdyka-Lojasiewicz property.
A neural doubly robust proxy causal learning framework using mean embeddings for treatment bridges provides consistent estimators for causal dose-response functions under unobserved confounding for continuous and structured treatments.
EquiNO with Q-DEIM creates reduced-order physics-informed surrogates for 3D hyperelastic RVEs that enforce equilibrium and periodicity by construction, achieve 10^3 speedups, and accurately interpolate and extrapolate stresses from few snapshots.
A partially deterministic Bernoulli sampling scheme for unitary-matrix compressed sensing that improves sample complexity and adds denoising guarantees over fully random methods.
QuantumXCT learns parameterized quantum circuits to model interaction-induced unitary transformations between non-interacting and interacting cellular state distributions from transcriptomic profiles.
A PINN pretrained on mechanistic synthetic data and fine-tuned experimentally is deployed in an EKF-style filter to estimate separator phase heights from flow rates alone.
VarQEC uses a distinguishability loss as a machine-learning objective to variationally discover resource-efficient encoding circuits optimized for given noise models.
An ensemble-variational method with 20 perturbed simulations per iteration recovers adjoint-quality open-loop forcing for a 2D open-cavity flow and partially reduces chaotic fluctuations.
Clockless FPGA circuits produce autonomous spiking neuron networks that achieve competitive audio classification accuracy with significantly lower power than conventional digital implementations.
A 3D mechano-geometric multicellular model integrates cell mechanics, irreversible wall growth, and deformable geometry to simulate apical stem-cell-driven plant morphogenesis.
Bayesian PINNs with Hamiltonian Monte Carlo sampling deliver the most consistent uncertainty estimates for turbulent flow inverse problems, while repulsive deep ensembles provide a faster but slightly less calibrated alternative.
Rothe's method stabilizes Gaussian wavepacket propagation for quantum dynamics, yielding grid-comparable accuracy for electronic and rovibrational processes including high-harmonic generation using remarkably few functions.
PINNs and DeepONets solve Newtonian plane Couette flow with dynamic wall slip; DeepONet achieves 0.36% mean relative error on unseen cases and 540X speedup over numerical methods.
Optimizes banana coils for tokamak-stellarator hybrids under force constraints and quantifies resulting geometries, fields, and force-reduction mechanisms.
PEPSKit.jl is a Julia package that supplies high-level algorithms for ground-state, time-evolution and finite-temperature iPEPS simulations with symmetry support on various lattices.
The authors convert classical software bug detection into quantum optimization instances and test QAOA, Grover, and QSVT on small examples for potential polynomial speedup.
MaRDI Open Interfaces supplies common interfaces for nonlinear optimization solvers, shown via an application to physics-informed neural network training on the viscous Burgers' equation.
This survey organizes LLM optimizer literature into categories and argues the field is shifting toward rigorous, multi-factor comparisons of convergence, memory, stability, and complexity.
citing papers explorer
-
ffortissimo: A Freeform Forward-Modeling Pipeline for High-Contrast Images of Circumstellar Disks Based on Automatic Differentiation
ffortissimo is a JAX-based freeform forward-modeling pipeline that fits complex dust distributions and infers scattering properties in KLIP-reduced images of circumstellar disks such as HR 4796A.
-
Curvature-aware dynamic precision approach for physics-informed neural networks
The curvature-aware precision controller adapts between FP32 and FP64 during PINN training to match double-precision accuracy at reduced computational cost.
-
Direct From Darwin: Deriving Advanced Optimizers From Evolutionary First Principles
SGD, approximations of Newton's method, natural gradient descent, and Adam are proven compatible with evolutionary dynamics when augmented with DLS noise, turning them into valid in silico simulations of asexual Darwinian evolution.
-
glmSTARMA -- An R-Package for fitting autoregressive spatio-temporal models following generalized linear models
An R package that unifies spatio-temporal STARMA and GARCH-type models through double generalized linear models, allowing simultaneous modeling of means and dispersion for count or continuous spatial time series.
-
A Physics-Informed B-Spline Framework for Continuous Approximation of Flow Data
PI-MFA optimizes tensor-product B-spline control points to balance data fidelity against PDE residuals, producing physically consistent continuous flow fields.
-
Reasoning Arena: Trace Tournaments When Verifiable Rewards Fall Short
Reasoning Arena converts non-diverse reward groups in RLVR into relative rewards via adaptive trace tournaments and Bradley-Terry fitting on anchor comparisons, claiming 7.6% average gains and 27-41% faster training on math/coding benchmarks.
-
Proximal Limited-Memory Quasi-Newton Methods for Nonsmooth Nonconvex Optimization
A proximal limited-memory quasi-Newton scheme is developed for nonsmooth nonconvex optimization, with global convergence proven under mild assumptions and rates under the Kurdyka-Lojasiewicz property.
-
Doubly Robust Proxy Causal Learning with Neural Mean Embeddings
A neural doubly robust proxy causal learning framework using mean embeddings for treatment bridges provides consistent estimators for causal dose-response functions under unobserved confounding for continuous and structured treatments.
-
Physics-Informed Reduced-Order Operator Learning for Hyperelasticity in Continuum Micromechanics
EquiNO with Q-DEIM creates reduced-order physics-informed surrogates for 3D hyperelastic RVEs that enforce equilibrium and periodicity by construction, achieve 10^3 speedups, and accurately interpolate and extrapolate stresses from few snapshots.
-
Partially deterministic sampling for compressed sensing with denoising guarantees
A partially deterministic Bernoulli sampling scheme for unitary-matrix compressed sensing that improves sample complexity and adds denoising guarantees over fully random methods.
-
QuantumXCT: Learning Interaction-Induced State Transformation in Cell-Cell Communication via Quantum Entanglement and Generative Modeling
QuantumXCT learns parameterized quantum circuits to model interaction-induced unitary transformations between non-interacting and interacting cellular state distributions from transcriptomic profiles.
-
Estimating Dense-Packed Zone Height in Liquid-Liquid Separation: A Physics-Informed Neural Network Approach
A PINN pretrained on mechanistic synthetic data and fine-tuned experimentally is deployed in an EKF-style filter to estimate separator phase heights from flow rates alone.
-
Learning Encodings by Maximizing State Distinguishability: Variational Quantum Error Correction
VarQEC uses a distinguishability loss as a machine-learning objective to variationally discover resource-efficient encoding circuits optimized for given noise models.
-
An ensemble-variational approach for open-loop flow control
An ensemble-variational method with 20 perturbed simulations per iteration recovers adjoint-quality open-loop forcing for a 2D open-cavity flow and partially reduces chaotic fluctuations.
-
Scalable neuromorphic computing from autonomous spiking dynamics in a clockless reconfigurable chip
Clockless FPGA circuits produce autonomous spiking neuron networks that achieve competitive audio classification accuracy with significantly lower power than conventional digital implementations.
-
3D mechano-geometric multicellular model of apical stem cell-driven plant morphogenesis
A 3D mechano-geometric multicellular model integrates cell mechanics, irreversible wall growth, and deformable geometry to simulate apical stem-cell-driven plant morphogenesis.
-
Uncertainty Quantification in PINNs for Turbulent Flows: Bayesian Inference and Repulsive Ensembles
Bayesian PINNs with Hamiltonian Monte Carlo sampling deliver the most consistent uncertainty estimates for turbulent flow inverse problems, while repulsive deep ensembles provide a faster but slightly less calibrated alternative.
-
Rothe's Method for Quantum Dynamics in Atoms and Molecules with Gaussian Wavepackets
Rothe's method stabilizes Gaussian wavepacket propagation for quantum dynamics, yielding grid-comparable accuracy for electronic and rovibrational processes including high-harmonic generation using remarkably few functions.
-
Solution of the Newtonian plane Couette flow with dynamic wall slip using machine-learning methods
PINNs and DeepONets solve Newtonian plane Couette flow with dynamic wall slip; DeepONet achieves 0.36% mean relative error on unseen cases and 540X speedup over numerical methods.
-
Understanding and Quantifying Banana Coil Magnetic Fields and Forces for Enhanced Optimisation
Optimizes banana coils for tokamak-stellarator hybrids under force constraints and quantifies resulting geometries, fields, and force-reduction mechanisms.
-
PEPSKit.jl: A Julia package for projected entangled-pair state simulations
PEPSKit.jl is a Julia package that supplies high-level algorithms for ground-state, time-evolution and finite-temperature iPEPS simulations with symmetry support on various lattices.
-
Towards Classical Software Verification using Quantum Computers
The authors convert classical software bug detection into quantum optimization instances and test QAOA, Grover, and QSVT on small examples for potential polynomial speedup.
-
Software package MaRDI Open Interfaces for improved interoperability in numerical optimization
MaRDI Open Interfaces supplies common interfaces for nonlinear optimization solvers, shown via an application to physics-informed neural network training on the viscous Burgers' equation.
-
Navigating LLM Valley: From AdamW to Memory-Efficient and Matrix-Based Optimizers
This survey organizes LLM optimizer literature into categories and argues the field is shifting toward rigorous, multi-factor comparisons of convergence, memory, stability, and complexity.
-
Quantum Machine Learning for Colorectal Cancer Data: Anastomotic Leak Classification and Risk Factors
Quantum neural networks achieve 83.3% sensitivity for anastomotic leak classification versus 66.7% for classical baselines on 14% prevalence clinical data.
-
A Tutorial Review of Bayesian Optimization with Gaussian Processes to Accelerate Stationary Point Searches
Bayesian optimization with Gaussian processes unifies minimization, single-point saddle searches, and double-ended path searches on potential energy surfaces through a shared six-step surrogate loop using derivative observations and inverse-distance kernels.
- Performance-Carbon Trade-Offs across Architectural Biases in Shear Flow Forecasting