Method makes non-interpretable models more intervenable.
problem Making non-interpretable models more understandable and controllable.
method Intervenability formalization and fine-tuning of black-box models.
result Fine-tuned black-box models are more intervenable and often better-calibrated.
New framework quantifies and reduces concept-based models' leakage.
problem Information leakage in concept-based models reduces interpretability.
method Information-theoretic framework with CTL and ICL measures.
result Measures predict model behaviour and identify leakage causes.
Concept-driven OPE reduces variance in off-policy decision evaluation.
problem High variance in off-policy decision evaluation due to limited sample sizes.
method Integrating human-explainable concepts into OPE to reduce variance.
result Concept-based OPE estimators remain unbiased and reduce variance when concepts are known and predefined.
New models handle survival analysis with concept-based learning.
problem Survival analysis tasks involving event times with censored data.
method SurvCBM and SurvRCM models integrating concept-based learning with survival analysis.
result SurvCBM outperforms traditional survival models in numerical experiments.
LCBM model improves image classification without human supervision.
problem Improving interpretability and generalization of unsupervised concept-based models.
method LCBM models concepts as random variables in a Bernoulli latent space, reducing the number of concepts without sacrificing performance.
result LCBM outperforms existing models in generalization and interpretability.
DCR improves interpretability of concept-based models by using neural networks to build rule structures.
problem Inability of concept-based models to provide transparent decision processes.
method DCR uses neural networks to build syntactic rule structures using concept embeddings and executes these rules on concept truth degrees.
result DCR improves interpretability by up to 25% on challenging benchmarks and discovers meaningful logic rules.
A method for concept-based learning using probabilistic inference and expert rules.
problem Concept-based learning with limited training data.
method Divide images into patches, transform into embeddings, cluster, and use frequentist inference to find concepts.
result FI-CBL outperforms concept bottleneck model in small data scenarios.
Debias concept-based explanations by removing confounding information.
problem Correlation between concepts and confounding features.
method Causal prior graph and two-stage regression technique.
result Success in removing biases and improving concept ranking.
ECBMs unify concept-based interpretations in deep learning models.
problem Suboptimal final accuracy and lack of concept interaction and conditional dependencies.
method ECBMs use a set of neural networks to define joint energy, enabling concept correction and conditional dependency quantification.
result ECBMs achieve higher accuracy and richer concept interpretations compared to state-of-the-art methods.
CB-SLICE identifies concept-based error slices in deep learning models.
problem Systematic errors in deep learning models on specific groups.
method Concept Bottleneck Models (CBMs) and concept representations.
result CB-SLICE outperforms state-of-the-art methods in error slice identification.
Proposes new methods for interpreting document classification models.
problem Interpretation fragility of attention-based neural networks.
method Corpus-level and concept-based explanation methods using attention weights.
result Extracts semantically meaningful keywords and concepts for model predictions.
Combines neural networks and expert rules for concept-based learning.
problem Extending concept-based learning with machine learning models.
method Form constraints for joint probability distribution and represent feasible set as a convex polytope.
result Neural networks can be trained to satisfy expert rules without violating them.
Interpretability has become an important topic of research as more machine learning (ML) models are deployed and widely used to make important decisions. Most of the current explanation methods provide explanations through feature importance scores, which identify features that are important for each individual input. …
MCD offers a complete model understanding for high-stake decisions.
problem Local model understanding in XAI methods is not sufficient for high-stake decisions.
method MCD extends concept-based methods to ensure global model understanding via multi-dimensional subspaces.
result MCD provides a complete model understanding, ensuring the model reasoning is related to the actual model.
This work abstracts deep neural networks into concept graphs for better interpretability in medical tasks.
problem Lack of interpretability in deep learning models, especially in medical domains.
method Developed a graphical representation of medical image processing models to understand concept-based reasoning.
result Extracted a concept-level graph that reveals the decision-making process of deep learning models.
Human explanations of high-level decisions are often expressed in terms of key concepts the decisions are based on. In this paper, we study such concept-based explainability for Deep Neural Networks (DNNs). First, we define the notion of completeness, which quantifies how sufficient a particular set of concepts is in e…
The problem of detecting terms that can be interesting to the advertiser is considered. If a company has already bought some advertising terms which describe certain services, it is reasonable to find out the terms bought by competing companies. A part of them can be recommended as future advertising terms to the compa…
Optimizes explanations for better listener understanding.
problem Insufficient consideration of listener preferences in concept-based explanations.
method Iterative training procedure based on direct preference optimization.
result Pragmatic explanations improve both model accuracy and user understanding.
Researchers apply concept-based explainability to EEG data.
problem Understanding the internal states of complex EEG transformer models.
method Concept Activation Vectors (CAVs) adapted for EEG data, using externally labeled datasets and anatomically defined concepts.
result Both approaches to concept formation yield valuable insights into EEG model representations.
Estimates joint causal effects using single-variable interventions on nonlinear models.
problem Estimating joint causal effects from single-variable interventions.
method Identifiability result and practical estimator for decomposing causal effects.
result Joint effects can be inferred without joint interventional data for nonlinear additive models.
This is a survey on old and new results as well as an introduction to various related basic notions and concepts, based on two talks given at the International Workshop on Geometry and Analysis in Kemerovo (Sobolev Institute of Mathematics, Kemerovo State University) and at the University of Krasnojarsk in June 2011. W…
This paper tackles CRL for multi-node interventions, achieving identifiability guarantees.
problem CRL under unknown multi-node interventions, focusing on single-node assumptions.
method Establishes identifiability results for general latent causal models under stochastic interventions.
result Identifiability up to ancestors using soft interventions, perfect identifiability using hard interventions.
This work tackles causal graph discovery with stochastic interventions to minimize the number of interventions.
problem Discovering the true causal graph from observational data with limited interventions.
method Proposes a stochastic intervention model and studies verification and search problems with approximation algorithms.
result Provides approximation algorithms with competitive ratios for verification and search problems.
Causal diagrams based on do intervention are useful tools to formalize, process and understand causal relationship among variables. However, the do intervention has controversial interpretation of causal questions for non-manipulable variables, and it also lacks the power to check the conditions related to counterfactu…
Interpretable ML methods for better decision-making with explanations.
problem Lack of transparency in black-box ML models.
method Use of Formal Concept Analysis and cooperative game theory to assess attribute importance and reduce attribute count.
result Developed methods to assess attribute importance and reduce attribute count in ML models.
Our goal is to identify beneficial interventions from observational data. We consider interventions that are narrowly focused (impacting few covariates) and may be tailored to each individual or globally enacted over a population. For applications where harmful intervention is drastically worse than proposing no change…
A concise review of recent few-shot meta-learning methods.
problem Mimicking human fast adaptation to new concepts based on prior knowledge.
method Categorized into four branches based on technical characteristics.
result Current challenges and future prospects identified.
Paper proposes scalable algorithm to estimate intervention targets in linear models.
problem Estimating intervention targets in linear models from observational and interventional data.
method The paper proposes a scalable algorithm that estimates intervention sites from the difference between precision matrices of observational and interventional datasets.
result The algorithm consistently identifies all intervention targets and updates observational Markov equivalence classes to interventional ones.
The intuition of risk is based on two main concepts: loss and variability. In this paper, we present a composition of risk and deviation measures, which contemplate these two concepts. Based on the proposed Limitedness axiom, we prove that this resulting composition, based on properties of the two components, is a cohe…
New method disentangles mixed interventional and observational data in SEMs.
problem Learning causal relationships from mixed interventional and observational data.
method Developed a method to disentangle mixed interventional and observational data in linear SEMs with Gaussian noise.
result The method can identify causal graphs up to their interventional Markov Equivalence Class.
Geometric framework detects concept frustration between human concepts and machine representations.
problem Aligning human concepts with machine learning representations.
method Geometric framework and similarity measures for detecting concept frustration.
result Concept frustration affects machine learning model performance and reorganizes learned concept representations.
The paper presents a method to estimate joint interventional distributions from marginal interventional data.
problem Estimating joint interventional distributions from marginal interventional data.
method The paper extends the Causal Maximum Entropy method to use interventional data and employs Lagrange duality to prove the solution lies in the exponential family.
result The method allows for causal feature selection and inference of joint interventional distributions.
In an attempt to gather a deeper understanding of how convolutional neural networks (CNNs) reason about human-understandable concepts, we present a method to infer labeled concept data from hidden layer activations and interpret the concepts through a shallow decision tree. The decision tree can provide information abo…
Bayesian method for causal discovery from unknown general interventions.
problem Learning causal DAGs from unknown interventions that modify parent sets.
method Bayesian approach with MCMC for approximating posterior DAGs and intervention targets.
result Bayesian method can identify DAGs and intervention targets up to equivalence classes.
Interventional data helps identify latent factors without distributional assumptions.
problem Identifying latent factors from interventional data without distributional assumptions.
method Leveraging geometric signatures of latent factors' support from interventional data.
result Latent causal factors can be identified up to permutation and scaling given data from perfect do-interventions.
IntDC framework uncovers causal relationships from non-interventional data.
problem Detecting causal relationships in non-interventional complex systems.
method Interventional Embedding Entropy (IEE) for causal strength measurement.
result IEE accurately finds causal edges and quantifies causal strength robustly.
Algorithm detects causal change points quickly with adaptive interventions.
problem Detecting changes in causal models with interventions.
method Centralization technique, Kullback-Leibler divergence for intervention selection, adaptive intervention policy.
result Theoretical first-order optimality and validation through simulations and real-world studies.
Method learns causal effects from multiple interventions in presence of unobserved confounders.
problem Disentangling causal effects from sets of interventions in the presence of unobserved confounders.
method Non-linear structural causal models with additive, multivariate Gaussian noise; algorithm that learns causal model parameters by pooling data from different regimes and maximizing combined likelihood.
result Identification proofs demonstrate that causal effects of single interventions can be learned from sets of interventions, even with unobserved confounders.
New algorithms improve causal graph discovery with adaptive interventions, even under worst-case interventional costs.
problem Discover causal relationships from data with adaptive interventions and node-dependent costs.
method Define new benchmarks and provide adaptive search algorithms for causal graph discovery.
result Logarithmic approximations achieved under various settings: atomic, bounded size interventions and generalized cost objectives.
We address the problem of optimal Central Bank intervention in the exchange rate market when interventions create feedback in the rate dynamics. In particular, we extend the work done on optimal impulse control by Cadenillas and Zapatero to incorporate temporary market reactions, of random duration and level, to Bank i…
GIT uses gradient estimators to target interventions for causal discovery.
problem Challenges in inferring causal structure from observational data.
method GIT uses gradient estimators to target interventions for causal discovery.
result GIT performs on par with competitive baselines, surpassing them in low-data regimes.
Paper relaxes faithfulness assumption for causal discovery using interventions.
problem Violation of faithfulness assumption in natural systems leads to incorrect causal structure identification.
method Use intervention-immediacy faithfulness assumption to identify causal structures with hard interventions.
result Interventions contain information about causal structure that can identify causal structures when faithfulness is violated.
FMI uses matching to mimic interventions for causal feature learning.
problem Challenges in causal discovery from observational data.
method Feature Matching Intervention (FMI) using matching to emulate perfect interventions.
result FMI outperforms in identifying causal features from observational data.
BaCaDI discovers causal structures from unknown interventions.
problem Inferring causal structures from unknown interventions with limited data.
method Bayesian framework with gradient-based variational inference.
result BaCaDI outperforms related methods in identifying causal structures and intervention targets.
Graph-coupled causal Bayesian optimization transfers information across related interventions.
problem Optimizing expensive systems where interventions are costly and causal effects are confounded.
method Ties intervention effects together through shared causal parameters, improving estimation.
result Information-gain and regret bounds show improved performance with shared mechanisms.
The paper tackles matching a desired mean in causal systems through shift interventions.
problem Matching a desired mean in causal systems.
method Defining Markov equivalence classes, proposing active learning strategies, deriving lower bounds.
result Proposed active learning strategies require fewer interventions than previous approaches, especially for certain graph classes.
Intervention loss stabilizes GAN training and reduces mode collapse.
problem Stabilizing GAN training and mitigating mode collapse.
method Intervention loss regularization term introduced into GAN objective.
result Improved GAN training stability and reduced mode collapse.
Model identifies causal structure from paired observational and interventional data with unknown soft interventions.
problem Identifying causal structure from observational and interventional data with unknown soft interventions.
method Proposes a scalable causal discovery model that aggregates subset-level PDAGs and applies contrastive cross-regime orientation rules.
result The model asymptotically recovers the identifiable PDAG and can orient additional edges compared to non-contrastive subset-restricted methods.