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.
hub Theoretical anchor
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
Theoretical anchor. 60% of citing Pith papers extend or build on this work's framework.
34
Pith papers citing it
Extended
60% of classified citations
hub tools
citation-role summary
background 5
citation-polarity summary
fields
cs.LG 16 cs.AI 6 cs.CV 2 cs.CY 2 stat.ML 2 cond-mat.mtrl-sci 1 cs.HC 1 cs.NE 1 cs.SE 1 physics.ao-ph 1roles
background 5representative citing papers
A Lie-algebraic kernel reparameterizes 3D rotationally anisotropic Gaussian processes with explicit principal length-scales and SO(3) orientations, matching full SPD flexibility but improving interpretability over axis-aligned ARD.
ProtoSSL discovers generalizable prototypes from unlabeled time-series via self-supervision and assigns them to new tasks for interpretable predictions, outperforming supervised baselines in low-data regimes on ECG datasets.
In-context symbolic regression methods improve robustness of symbolic formula recovery from KANs, cutting median OFAT test MSE by up to 99.8 percent across hyperparameter sweeps.
Chain-of-Thought reasoning in LLMs is often unfaithful, with models relying on it variably by task and less so as models scale larger.
MF-Net learns a shared field state and mechanical transition rule from trajectories to deliver competitive forecasting and recoverable relation matrices on Lorenz-96 and real systems.
Merging real-valued GOMEA with GP-GOMEA enables simultaneous optimization of constants and expression structure, generally outperforming other constant-handling techniques in symbolic regression.
TSL learns separable rank-1 univariate function products via stagewise greedy fitting with orthogonal refitting, reconstructible from first-order partial dependence plots with approximation guarantees.
FlagGAM builds sparse univariate rule bases from features and feeds them into a restricted additive model, achieving competitive accuracy with superior robustness to missingness and noise on tabular benchmarks.
DEM distills XGBoost into a residual decision tree with a new fidelity metric for interpretable anomaly detection in WBAN data, reporting AUC 0.9964 and 0.9047 with 0.17ms inference.
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
A think-aloud study reveals that AI tools in early research misrepresent uncertainty, obscure provenance, and create fragile trust, leading researchers to develop compensatory strategies to preserve scholarly judgment.
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.
ZKMLOps is an MLOps framework that uses zero-knowledge proofs to generate verifiable cryptographic evidence of AI model compliance without revealing confidential information.
Introduces a unified evaluation framework for XAI using five principled metrics and the PGCA method that fuses grid perturbation with Grad-CAM++ , reporting top scores in fidelity, interpretability and fairness on ResNet-50 models across five image domains.
MSI is a multimodal representation learning framework that identifies key microstructural features governing mechanical behavior in structural alloys from spatial observations.
Adds a trainable feature selection layer to NAM and NBM to cut computational cost, enable two-input interaction networks in high dimensions, and match or exceed state-of-the-art GAM performance.
VP2O maps PPO to SVGD in a MoE architecture using functional kernels and expert orthogonalization, claiming +179 ELO on Codeforces and 32% token reduction on AIME for a 33B/4B model.
An AI recommender system improves Cox Proportional Hazards model performance for predicting patient falls by suggesting 23 feature exclusions, 2 non-linear terms, and 221 interactions, raising C-index from 0.805 to 0.815.
ProtoPathway fuses prototype-based histopathology encoding with pathway-aware graph neural networks for multimodal cancer survival prediction and native biological attribution.
AC-GATE is a lag-gated neural encoder that conditions lag-weight distributions on entity proxies to recover heterogeneous lags as structural model outputs in panel time series.
CCSS-IX is a context-conditioned structured simulator for wastewater digital twins that uses adaptive expert mixing and self-falsifying conformal decision rules to reduce unsafe actions while maintaining low prediction error on real plant and benchmark data.
Soft Learning optimally combines heterogeneous ML specialists via cross-validated non-negative least squares, achieving top performance on 70% of 37 datasets with formal guarantees and 72-435x CPU speedups over deep networks.
A novel algorithm learns sets of optimal quantile regression trees to predict full conditional distributions interpretably and efficiently.
citing papers explorer
-
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.
-
Interpretable Machine Learning for Spatial Science: A Lie-Algebraic Kernel for Rotationally Anisotropic Gaussian Processes
A Lie-algebraic kernel reparameterizes 3D rotationally anisotropic Gaussian processes with explicit principal length-scales and SO(3) orientations, matching full SPD flexibility but improving interpretability over axis-aligned ARD.
-
ProtoSSL: Interpretable Prototype Learning from Unlabeled Time-Series Data
ProtoSSL discovers generalizable prototypes from unlabeled time-series via self-supervision and assigns them to new tasks for interpretable predictions, outperforming supervised baselines in low-data regimes on ECG datasets.
-
In-Context Symbolic Regression for Robustness-Improved Kolmogorov-Arnold Networks
In-context symbolic regression methods improve robustness of symbolic formula recovery from KANs, cutting median OFAT test MSE by up to 99.8 percent across hyperparameter sweeps.
-
Measuring Faithfulness in Chain-of-Thought Reasoning
Chain-of-Thought reasoning in LLMs is often unfaithful, with models relying on it variably by task and less so as models scale larger.
-
Mechanical Field Networks: Structured Neural Dynamics for Multivariate Systems
MF-Net learns a shared field state and mechanical transition rule from trajectories to deliver competitive forecasting and recoverable relation matrices on Lorenz-96 and real systems.
-
Simultaneous Model-Based Evolution of Constants and Expression Structure in GP-GOMEA for Symbolic Regression
Merging real-valued GOMEA with GP-GOMEA enables simultaneous optimization of constants and expression structure, generally outperforming other constant-handling techniques in symbolic regression.
-
Beyond Additive Decompositions: Interpretability Through Separability
TSL learns separable rank-1 univariate function products via stagewise greedy fitting with orthogonal refitting, reconstructible from first-order partial dependence plots with approximation guarantees.
-
FlagGAM: Rule-Basis Generalized Additive Models for Explainable Tabular Prediction
FlagGAM builds sparse univariate rule bases from features and feeds them into a restricted additive model, achieving competitive accuracy with superior robustness to missingness and noise on tabular benchmarks.
-
DEM: A Distilled Explanation Model for Interpretable Anomaly Detection in Physiological Sensor Networks
DEM distills XGBoost into a residual decision tree with a new fidelity metric for interpretable anomaly detection in WBAN data, reporting AUC 0.9964 and 0.9047 with 0.17ms inference.
-
Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
-
How Researchers Navigate Accountability, Transparency, and Trust When Using AI Tools in Early-Stage Research: A Think-Aloud Study
A think-aloud study reveals that AI tools in early research misrepresent uncertainty, obscure provenance, and create fragile trust, leading researchers to develop compensatory strategies to preserve scholarly judgment.
-
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.
-
"Show Me You Comply... Without Showing Me Anything": Zero-Knowledge Software Auditing for AI-Enabled Systems
ZKMLOps is an MLOps framework that uses zero-knowledge proofs to generate verifiable cryptographic evidence of AI model compliance without revealing confidential information.
-
A Unified Framework for Evaluating and Enhancing the Transparency of Explainable AI Methods via Perturbation-Gradient Consensus Attribution
Introduces a unified evaluation framework for XAI using five principled metrics and the PGCA method that fuses grid perturbation with Grad-CAM++ , reporting top scores in fidelity, interpretability and fairness on ResNet-50 models across five image domains.
-
Interpretable Material Spatial Intelligence for Discovery of Governing Microstructural Features
MSI is a multimodal representation learning framework that identifies key microstructural features governing mechanical behavior in structural alloys from spatial observations.
-
Neural Additive and Basis Models with Feature Selection and Interactions
Adds a trainable feature selection layer to NAM and NBM to cut computational cost, enable two-input interaction networks in high dimensions, and match or exceed state-of-the-art GAM performance.
-
Variational Proximal Policy Optimization
VP2O maps PPO to SVGD in a MoE architecture using functional kernels and expert orthogonalization, claiming +179 ELO on Codeforces and 32% token reduction on AIME for a 33B/4B model.
-
Explainable AI for Data-Driven Design of High-Dimensional Predictive Studies
An AI recommender system improves Cox Proportional Hazards model performance for predicting patient falls by suggesting 23 feature exclusions, 2 non-linear terms, and 221 interactions, raising C-index from 0.805 to 0.815.
-
ProtoPathway: Biologically Structured Prototype-Pathway Fusion for Multimodal Cancer Survival Prediction
ProtoPathway fuses prototype-based histopathology encoding with pathway-aware graph neural networks for multimodal cancer survival prediction and native biological attribution.
-
Discovering Entity-Conditioned Lag Heterogeneity: A Lag-Gated Neural Audit Framework for Panel Time Series
AC-GATE is a lag-gated neural encoder that conditions lag-weight distributions on entity proxies to recover heterogeneous lags as structural model outputs in panel time series.
-
Explainable Wastewater Digital Twins: Adaptive Context-Conditioned Structured Simulators with Self-Falsifying Decision Support
CCSS-IX is a context-conditioned structured simulator for wastewater digital twins that uses adaptive expert mixing and self-falsifying conformal decision rules to reduce unsafe actions while maintaining low prediction error on real plant and benchmark data.
-
Soft Learning
Soft Learning optimally combines heterogeneous ML specialists via cross-validated non-negative least squares, achieving top performance on 70% of 37 datasets with formal guarantees and 72-435x CPU speedups over deep networks.
-
Interpretable Quantile Regression by Optimal Decision Trees
A novel algorithm learns sets of optimal quantile regression trees to predict full conditional distributions interpretably and efficiently.
-
From Large Language Model Predicates to Logic Tensor Networks: Neurosymbolic Offer Validation in Regulated Procurement
A neurosymbolic pipeline extracts predicates from offer texts with an LLM and validates them via Logic Tensor Networks, delivering performance comparable to standard models plus built-in interpretability on a real corpus.
-
Interpretable Policy Distillation for Power Grid Topology Control
PPO policy for grid topology control is distilled into decision trees and random forests that outperform the teacher on reward and survival time with lower inference cost and high interpretability.
-
AIMBio-Mat: An AI-Native FAIR Platform for Closed-Loop Materials Discovery and Biomedical Translation
AIMBio-Mat is a conceptual blueprint for an AI-native, FAIR, governance-aware decision layer that formulates biomedical-materials discovery as constrained multi-objective optimization under uncertainty.
-
Robust and Explainable Divide-and-Conquer Learning for Intrusion Detection
A divide-and-conquer method decomposes network intrusion detection into focused subtasks, allowing lightweight models to gain up to 43.3% higher local accuracy and 257x smaller size while improving robustness and explainability.
-
LLMs Should Not Yet Be Credited with Decision Explanation
LLMs support decision prediction and rationale generation but lack evidence for genuine decision explanation, requiring stricter standards to avoid over-crediting.
-
Human Agency, Causality, and the Human Computer Interface in High-Stakes Artificial Intelligence
The paper proposes a Causal-Agency Framework to restore human causal control at AI interfaces by combining causal models, uncertainty quantification, and human-centered evaluation.
-
Toward Aristotelian Medical Representations: Backpropagation-Free Layer-wise Analysis for Interpretable Generalized Metric Learning on MedMNIST
A-ROM delivers competitive MedMNIST performance via pretrained ViT metric spaces, a concept dictionary, and kNN without backpropagation or fine-tuning, framed as interpretable few-shot learning under the Platonic Representation Hypothesis.
-
AI to Learn 2.0: A Deliverable-Oriented Governance Framework and Maturity Rubric for Opaque AI in Learning-Intensive Domains
AI to Learn 2.0 is a deliverable-oriented framework with a seven-dimension maturity rubric and capability-evidence ladder that permits opaque AI for exploration but requires final outputs to be auditable, transferable, and supported by human-attributable evidence.
-
Comparing Post-Hoc Explainable AI Methods for Interpreting Black-Box EEG Models in Depression Detection
Gradient- and perturbation-based XAI methods show substantial agreement on frontal, temporal, and posterior EEG regions for an InceptionTime MDD classifier, while DeepSHAP differs, with overall partial convergence and method-dependent variability.
- Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions