An argument paper reframes LLM explainability as an embodied, situated practice based on Dourish and enactivist cognition, identifying ontological obstacles in internal explanations and advocating affordance-based designs.
super hub Canonical reference
Deep learning
Canonical reference. 100% of citing Pith papers cite this work as background.
74
Pith papers citing it
71.9k
external citations · Crossref
Background
100% of classified citations
hub tools
citation-role summary
background 14
method 1
citation-polarity summary
authors
co-cited works
fields
cs.LG 26 cs.CV 8 cs.AI 3 cs.CL 3 eess.SP 3 physics.ao-ph 3 quant-ph 3 astro-ph.IM 2 astro-ph.SR 2 math.OC 2polarities
background 15representative citing papers
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.
eCNNTO applies an element-wise CNN with residual connections and final-stage training data to accelerate density-based topology optimization while generalizing across boundary conditions, loads, geometries, and mesh sizes.
Optimizing training data via a differentiable SCM yields climate emulators that outperform those trained on six standard ScenarioMIP pathways while using less data and isolating distinct forcing responses.
Spatial multiplexing in optical neural networks is repurposed as a trainable representational coordinate, demonstrated in multi-layer architectures for image classification, regression, and hybrid vision-language captioning with over one million optical phase parameters.
A U-Net surrogate with multigroup attention pooling is trained on OpenMC sensitivity data and combined with gradient optimization to generate grid-based critical experiment geometries that achieve c_k values up to 0.97757 for HALEU fuel validation.
CAML meta-learns a progressively refined inductive bias from active-learning queries to improve robustness to spurious correlations, reporting accuracy gains on minority groups across several benchmarks.
In a combinatorial toy setting, winning lottery tickets preserve families of compatible feature locations in early feature space that balance proximity to final codes with low interference, rather than specific weight subnetworks.
DualTCN is the first deep-learning model for time-domain marine CSEM inversion that regresses four earth parameters, achieves high accuracy on simulated data, and runs up to 21,000 times faster than classical optimizers.
Broximal Alignment is a novel condition under which the Ball Proximal Point Method converges to global minima in non-convex settings, generalizing quasiconvexity, star convexity, and related frameworks.
Neural decompositionality is defined via decision-boundary semantic preservation, and language transformers largely satisfy it under SAVED while vision models often do not.
Quantum circuits for coherent multilayer neural network inference achieve quadratic to polylogarithmic speedups over classical methods depending on quantum data access models for inputs and weights.
A machine learning model called neural quantum propagator is introduced to efficiently solve non-Markovian quantum dynamics described by HEOM and applied to simulate spectra of the FMO complex.
Local 2- and 3-cycles enhance RNN computational capacity for Boolean functions, predicted by structural statistics, while adding interneurons boosts large networks.
Empirical audit of LAION-2B-en and LAION-2B-multi finds overrepresentation of young adults, White people, and males plus stereotypical emotion associations across two attribute classifiers.
Hybrid neural-process model derives biokinetic parameters from genomic traits for soil organic matter turnover, with ecological constraints, and outperforms baselines on synthetic and real data.
Introduces the directional linear separability measure (LSM) as an asymmetric diagnostic for one-sided affine separability of neural representations.
Tabular foundation models applied to PHM via signal-to-table conversion achieve the best average ranks across prognostic and diagnostic tasks and remain competitive in low-data regimes.
A parameter-efficient dual-encoder model with differentiable Choquet integral fusion improves underwater acoustic classification accuracy over single-encoder baselines on DeepShip and ShipsEar datasets.
Observation of ultrafast large dynamic magnetic soliton formation inside the linear spin-wave band in garnet films, extending tens of microns and collapsing into short-wavelength spin waves at large distances.
Pilot study uses pretrained video encoder features from lung ultrasound to predict 30-day CHF readmission, finding lower-lung views and temporal differences most informative with top MLP F1 of 0.80.
A Bayesian CNN maps 2D slitless spectral images to redshift estimates with NMAD precision 0.0104 for SNR_GI >=1 and better for brighter sources, while remaining robust to wavelength calibration errors via spatial augmentations.
Scaling vision models by depth and parameter count does not consistently improve localisation-based explanation quality across architectures, datasets, and post-hoc methods; smaller models often perform comparably or better.
Mixing real UAV imagery with 2101 AI-generated image-mask pairs improves semantic segmentation F1 scores for fine-grained forest species by over 15 percentage points overall and up to 30 points for rare classes.
citing papers explorer
-
Embodied Explainability and Ontological Obstacles: Why We Struggle to Explain the Answers of Large Language Models (LLMs)
An argument paper reframes LLM explainability as an embodied, situated practice based on Dourish and enactivist cognition, identifying ontological obstacles in internal explanations and advocating affordance-based designs.
-
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.
-
eCNNTO: A Highly Generalizable ConvNet for Accelerating Topology Optimization
eCNNTO applies an element-wise CNN with residual connections and final-stage training data to accelerate density-based topology optimization while generalizing across boundary conditions, loads, geometries, and mesh sizes.
-
Optimal scenario design for climate emulation
Optimizing training data via a differentiable SCM yields climate emulators that outperform those trained on six standard ScenarioMIP pathways while using less data and isolating distinct forcing responses.
-
Multi-channel Optical Vision Model
Spatial multiplexing in optical neural networks is repurposed as a trainable representational coordinate, demonstrated in multi-layer architectures for image classification, regression, and hybrid vision-language captioning with over one million optical phase parameters.
-
Inverse Critical Experiment Design via Gradient Optimization and a Multigroup Attention-Based Neural Network Architecture
A U-Net surrogate with multigroup attention pooling is trained on OpenMC sensitivity data and combined with gradient optimization to generate grid-based critical experiment geometries that achieve c_k values up to 0.97757 for HALEU fuel validation.
-
Cumulative Meta-Learning from Active Learning Queries for Robustness to Spurious Correlations
CAML meta-learns a progressively refined inductive bias from active-learning queries to improve robustness to spurious correlations, reporting accuracy gains on minority groups across several benchmarks.
-
Toy Combinatorial Interpretability Models Reveal Lottery Tickets in Early Feature Space
In a combinatorial toy setting, winning lottery tickets preserve families of compatible feature locations in early feature space that balance proximity to final codes with low interference, rather than specific weight subnetworks.
-
DualTCN: A Physics-Constrained Temporal Convolutional Network for 2 Time-Domain Marine CSEM Inversion
DualTCN is the first deep-learning model for time-domain marine CSEM inversion that regresses four earth parameters, achieves high accuracy on simulated data, and runs up to 21,000 times faster than classical optimizers.
-
Broximal Alignment for Global Non-Convex Optimization
Broximal Alignment is a novel condition under which the Ball Proximal Point Method converges to global minima in non-convex settings, generalizing quasiconvexity, star convexity, and related frameworks.
-
On the Decompositionality of Neural Networks
Neural decompositionality is defined via decision-boundary semantic preservation, and language transformers largely satisfy it under SAVED while vision models often do not.
-
Accelerating Inference for Multilayer Neural Networks with Quantum Computers
Quantum circuits for coherent multilayer neural network inference achieve quadratic to polylogarithmic speedups over classical methods depending on quantum data access models for inputs and weights.
-
Non-markovian neural quantum propagator and its application to the simulation of ultrafast nonlinear spectra
A machine learning model called neural quantum propagator is introduced to efficiently solve non-Markovian quantum dynamics described by HEOM and applied to simulate spectra of the FMO complex.
-
Identifying structural design principles shaping the computational abilities of recurrent neural networks
Local 2- and 3-cycles enhance RNN computational capacity for Boolean functions, predicted by structural statistics, while adding interneurons boosts large networks.
-
Unmasking LAION-5B: Age, Gender, Race, and Emotion Biases in Large-Scale Image Datasets
Empirical audit of LAION-2B-en and LAION-2B-multi finds overrepresentation of young adults, White people, and males plus stereotypical emotion associations across two attribute classifiers.
-
Constrained hybrid modelling to predict microbial dynamics and organic matter turnover in soil systems
Hybrid neural-process model derives biokinetic parameters from genomic traits for soil organic matter turnover, with ecological constraints, and outperforms baselines on synthetic and real data.
-
A Geometric Measure of Linear Separability for Neural Representations
Introduces the directional linear separability measure (LSM) as an asymmetric diagnostic for one-sided affine separability of neural representations.
-
Towards Unified and Data-Efficient Prognostics and Health Management with Tabular Foundation Models
Tabular foundation models applied to PHM via signal-to-table conversion achieve the best average ranks across prognostic and diagnostic tasks and remain competitive in low-data regimes.
-
Parameter-efficient Dual-encoder Architecture with Differentiable Choquet Integral Fusion for Underwater Acoustic Classification
A parameter-efficient dual-encoder model with differentiable Choquet integral fusion improves underwater acoustic classification accuracy over single-encoder baselines on DeepShip and ShipsEar datasets.
-
Ultrafast formation of a large dynamic magnetic soliton
Observation of ultrafast large dynamic magnetic soliton formation inside the linear spin-wave band in garnet films, extending tens of microns and collapsing into short-wavelength spin waves at large distances.
-
Prognostic Value of Lung Ultrasound Biomarkers for Readmission Risk in Congestive Heart Failure: A Pilot Data-Driven Analysis
Pilot study uses pretrained video encoder features from lung ultrasound to predict 30-day CHF readmission, finding lower-lung views and temporal differences most informative with top MLP F1 of 0.80.
-
Extracting redshifts from 2D slitless spectroscopic images using deep learning for the CSST galaxy survey
A Bayesian CNN maps 2D slitless spectral images to redshift estimates with NMAD precision 0.0104 for SNR_GI >=1 and better for brighter sources, while remaining robust to wavelength calibration errors via spatial augmentations.
-
Scaling Vision Models Does Not Consistently Improve Localisation-Based Explanation Quality
Scaling vision models by depth and parameter count does not consistently improve localisation-based explanation quality across architectures, datasets, and post-hoc methods; smaller models often perform comparably or better.
-
Leveraging Image Generators to Address Training Data Scarcity: The Gen4Regen Dataset for Forest Regeneration Mapping
Mixing real UAV imagery with 2101 AI-generated image-mask pairs improves semantic segmentation F1 scores for fine-grained forest species by over 15 percentage points overall and up to 30 points for rare classes.
-
A Meta Reinforcement Learning Approach to Goals-Based Wealth Management
MetaRL pre-trained on GBWM problems delivers near-optimal dynamic strategies in 0.01s achieving 97.8% of DP optimal utility and handles larger problems where DP fails.
-
Lottery BP: Unlocking Quantum Error Decoding at Scale
Lottery BP adds randomness to belief propagation decoding and uses syndrome voting to achieve far higher accuracy on topological quantum codes while reducing reliance on expensive global decoders.
-
Open-Vocabulary Semantic Segmentation Network Integrating Object-Level Label and Scene-Level Semantic Features for Multimodal Remote Sensing Images
TSMNet uses a dual-branch text encoder and text-guided fusion module to integrate scene-level semantic and object-level label features from text with visual embeddings, achieving superior open-vocabulary segmentation on new multimodal remote sensing datasets.
-
Mistake gating leads to energy and memory efficient continual learning
Mistake-gated plasticity reduces neural network updates by 50-80% by gating changes on classification errors, improving efficiency for continual learning without added hyperparameters.
-
Extraction of linearized models from pre-trained networks via knowledge distillation
Koopman theory plus knowledge distillation yields linearized models from pre-trained nets that outperform standard least-squares Koopman approximations on MNIST and Fashion-MNIST in accuracy and stability.
-
Neural Networks With Dense Weights Are Not Universal Approximators
Dense ReLU networks under natural weight and dimension constraints fail to approximate certain Lipschitz functions, unlike unrestricted networks.
-
iPDB -- Optimizing Semantic SQL Queries
iPDB adds a predict operator and semantic query optimizations to SQL so that LLM and ML calls run efficiently inside the database, delivering 2.5x average and up to 30x speedup over prior systems.
-
Fusion or Confusion? Multimodal Complexity Is Not All You Need
Complex multimodal architectures do not reliably outperform unimodal baselines or a simple multimodal baseline under standardized evaluation.
-
HOLE: Homological Observation of Latent Embeddings for Neural Network Interpretability
HOLE applies persistent homology to latent embeddings in neural networks and uses visualizations such as cluster flow diagrams to reveal patterns of class separation, feature disentanglement, and robustness.
-
Pulse Shape Discrimination Algorithms: Survey and Benchmark
A survey and benchmark of ~60 PSD algorithms on two radiation datasets finds deep learning models (MLPs and hybrids) often outperform traditional statistical methods, with an open-source Python/MATLAB toolbox and datasets released.
-
Relating Simple Sentence Representations in Deep Neural Networks and the Brain
BERT activations show strongest correlation with MEG data for simple sentences; DNN representations generate synthetic brain data that improves stimuli decoding accuracy.
-
SEAGAN: domain-Specific and Edge-Aware Graph Attention Network for Dynamic Plant Processes
SEAGAN applies a domain-specific graph attention network to classify limitation states in A-Ci curves, achieving F1-score 0.857 and accuracy 0.882 on synthetic data with known ground truth.
-
Dirichlet-Guided Group Forecasting for Alleviating Over-smoothing in Time Series Forecasting
DGF explicitly models multiple mode-conditioned predictive distributions via Dirichlet-guided sampling and reward optimization to preserve dynamical features in time series forecasts.
-
A Variability-Based Framework for Interpretable Naming in Formal and Relational Concept Analysis
Presents a configurable variability-based framework for LLM-assisted naming of formal concepts in FCA and RCA, illustrated on a small pizzeria relational dataset.
-
A Simulation Methodology Testbed for Typhoon Sensitivity Analysis: Framework Development and Perturbation-Response Experiments with the Pangu Weather Model
A MATLAB/ONNX testbed integrates the Pangu AI model with PID closed-loop control to perform single-input single-output perturbation-response experiments on typhoon track and intensity.
-
Classification of Single and Mixed Partial Discharges under Switching Voltage Using an AWA-CNN Framework
AWA patterns from PD pulse amplitude, width, and area enable CNNs to classify single and mixed partial discharge sources under switching voltage with over 96% test accuracy.
-
Generation of Heterogeneous PET Images from Uniform Organ Activity Maps Using a Pretrained Domain-Adapted Diffusion Model
A domain-adapted diffusion model synthesizes heterogeneous PET images from uniform organ activity maps, achieving high quantitative accuracy (CCC > 0.92) and visual realism comparable to real scans.
-
Foundation Models for Credit Risk Prediction: A Game Changer?
Tabular foundation models outperform standard methods in credit risk PD and LGD tasks, with larger gains on smaller datasets when used out-of-the-box.
-
Interpretable Neural Networks to Predict Momentum Fluxes of Orographic Gravity Waves
Neural networks predict orographic gravity wave momentum fluxes from coarse state variables with offline R² of 0.56-0.72, learn physically meaningful relationships via SHAP, and are compared to the Lott-Miller parameterization.
-
Machine Learning Enhanced Laser Spectroscopy for Multi-Species Gas Detection in Complex and Harsh Environments
Machine learning methods including denoising autoencoders, unsupervised interference mitigation, blind source separation, and certifiable classification are developed and experimentally validated to improve multi-species laser spectroscopy under complex conditions.
-
Predicting Associations between Solar Flares and Coronal Mass Ejections Using SDO/HMI Magnetograms and a Hybrid Neural Network
Hybrid neural network predicts eruptive versus confined solar flares from SDO/HMI magnetogram sequences, reports good performance, and links results to magnetic flux cancellation in polarity inversion lines.
-
Determination of Nanoparticle and Microdroplet Parameters in Levitating Microdroplets of Suspension by Speckle Image Analysis Using Convolutional Neural Networks
CNNs trained on speckle images from levitating TiO2 suspension microdroplets classify droplet diameter with better than 6% accuracy and provide useful discrimination for nanoparticle concentration and diameter, including simultaneous three-parameter classification.
-
Operator-Theoretic Energy Functionals for Impulse-Excited Nonstationary Signal Analysis
An operator-based Energy Concentration Index yields the IMRED detector that identifies defect-induced changes in impulse responses with AUC 0.908, outperforming standard Fourier and wavelet energy measures.
-
Vanishing Contributions: A Unified Framework for Smooth and Iterative Model Compression
VCON is a unified framework for smooth iterative DNN compression that uses parallel execution and an affine combination to progressively replace the original model with its compressed form during fine-tuning.
-
Bayesian Reasoning for Physics Informed Neural Networks
Introduces Laplace-approximated Bayesian PINNs for automatic loss-weight optimization when solving PDEs such as heat, wave, and Burgers equations.
-
General Inverse Design of Thin-Film Metamaterials With Convolutional Neural Networks
Convolutional neural networks are shown to perform inverse design of thin-film metamaterial stacks by learning the mapping from structure to ellipsometric and reflectance/transmittance spectra, with efficiency gains over traditional optimization as layer count increases.