This paper introduces KnowHalu, a novel approach for detecting hallucinations in text generated by large language models (LLMs), utilizing step-wise reasoning, multi-formulation query, multi-form knowledge for factual checking, and fusion-based detection mechanism. As LLMs are increasingly applied across various domains, ensuring that their outputs are not hallucinated is critical. Recognizing the limitations of existing approaches that either rely on the self-consistency check of LLMs or perform post-hoc fact-checking without considering the complexity of queries or the form of knowledge, KnowHalu proposes a two-phase process for hallucination detection. In the first phase, it identifies non-fabrication hallucinations--responses that, while factually correct, are irrelevant or non-specific to the query. The second phase, multi-form based factual checking, contains five key steps: reasoning and query decomposition, knowledge retrieval, knowledge optimization, judgment generation, and judgment aggregation. Our extensive evaluations demonstrate that KnowHalu significantly outperforms SOTA baselines in detecting hallucinations across diverse tasks, e.g., improving by 15.65% in QA tasks and 5.50% in summarization tasks, highlighting its effectiveness and versatility in detecting hallucinations in LLM-generated content.
Feature selection is a crucial step in building machine learning models. This process is often achieved with accuracy as an objective, and can be cumbersome and computationally expensive for large-scale datasets. Several additional model performance characteristics such as fairness and robustness are of importance for model development. As regulations are driving the need for more trustworthy models, deployed models need to be corrected for model characteristics associated with responsible artificial intelligence. When feature selection is done with respect to one model performance characteristic (eg. accuracy), feature selection with secondary model performance characteristics (eg. fairness and robustness) as objectives would require going through the computationally expensive selection process from scratch. In this paper, we introduce the problem of feature \emph{reselection}, so that features can be selected with respect to secondary model performance characteristics efficiently even after a feature selection process has been done with respect to a primary objective. To address this problem, we propose REFRESH, a method to reselect features so that additional constraints that are desirable towards model performance can be achieved without having to train several new models. REFRESH's underlying algorithm is a novel technique using SHAP values and correlation analysis that can approximate for the predictions of a model without having to train these models. Empirical evaluations on three datasets, including a large-scale loan defaulting dataset show that REFRESH can help find alternate models with better model characteristics efficiently. We also discuss the need for reselection and REFRESH based on regulation desiderata.
In the domain of Natural Language Processing (NLP), Large Language Models (LLMs) have demonstrated promise in text-generation tasks. However, their educational applications, particularly for domain-specific queries, remain underexplored. This study investigates LLMs' capabilities in educational scenarios, focusing on concept graph recovery and question-answering (QA). We assess LLMs' zero-shot performance in creating domain-specific concept graphs and introduce TutorQA, a new expert-verified NLP-focused benchmark for scientific graph reasoning and QA. TutorQA consists of five tasks with 500 QA pairs. To tackle TutorQA queries, we present CGLLM, a pipeline integrating concept graphs with LLMs for answering diverse questions. Our results indicate that LLMs' zero-shot concept graph recovery is competitive with supervised methods, showing an average 3% F1 score improvement. In TutorQA tasks, LLMs achieve up to 26% F1 score enhancement. Moreover, human evaluation and analysis show that CGLLM generates answers with more fine-grained concepts.
Although neural models have achieved remarkable performance, they still encounter doubts due to the intransparency. To this end, model prediction explanation is attracting more and more attentions. However, current methods rarely incorporate external knowledge and still suffer from three limitations: (1) Neglecting concept completeness. Merely selecting concepts may not sufficient for prediction. (2) Lacking concept fusion. Failure to merge semantically-equivalent concepts. (3) Difficult in manipulating model behavior. Lack of verification for explanation on original model. To address these issues, we propose a novel knowledge-aware neuron interpretation framework to explain model predictions for image scene classification. Specifically, for concept completeness, we present core concepts of a scene based on knowledge graph, ConceptNet, to gauge the completeness of concepts. Our method, incorporating complete concepts, effectively provides better prediction explanations compared to baselines. Furthermore, for concept fusion, we introduce a knowledge graph-based method known as Concept Filtering, which produces over 23% point gain on neuron behaviors for neuron interpretation. At last, we propose Model Manipulation, which aims to study whether the core concepts based on ConceptNet could be employed to manipulate model behavior. The results show that core concepts can effectively improve the performance of original model by over 26%.
Transformer-based models excel in various natural language processing (NLP) tasks, attracting countless efforts to explain their inner workings. Prior methods explain Transformers by focusing on the raw gradient and attention as token attribution scores, where non-relevant information is often considered during explanation computation, resulting in confusing results. In this work, we propose highlighting the important information and eliminating irrelevant information by a refined information flow on top of the layer-wise relevance propagation (LRP) method. Specifically, we consider identifying syntactic and positional heads as important attention heads and focus on the relevance obtained from these important heads. Experimental results demonstrate that irrelevant information does distort output attribution scores and then should be masked during explanation computation. Compared to eight baselines on both classification and question-answering datasets, our method consistently outperforms with over 3\% to 33\% improvement on explanation metrics, providing superior explanation performance. Our anonymous code repository is available at: https://github.com/LinxinS97/Mask-LRP
When individuals are subject to adverse outcomes from machine learning models, providing a recourse path to help achieve a positive outcome is desirable. Recent work has shown that counterfactual explanations - which can be used as a means of single-step recourse - are vulnerable to privacy issues, putting an individuals' privacy at risk. Providing a sequential multi-step path for recourse can amplify this risk. Furthermore, simply adding noise to recourse paths found from existing methods can impact the realism and actionability of the path for an end-user. In this work, we address privacy issues when generating realistic recourse paths based on instance-based counterfactual explanations, and provide PrivRecourse: an end-to-end privacy preserving pipeline that can provide realistic recourse paths. PrivRecourse uses differentially private (DP) clustering to represent non-overlapping subsets of the private dataset. These DP cluster centers are then used to generate recourse paths by forming a graph with cluster centers as the nodes, so that we can generate realistic - feasible and actionable - recourse paths. We empirically evaluate our approach on finance datasets and compare it to simply adding noise to data instances, and to using DP synthetic data, to generate the graph. We observe that PrivRecourse can provide paths that are private and realistic.
As systems based on opaque Artificial Intelligence (AI) continue to flourish in diverse real-world applications, understanding these black box models has become paramount. In response, Explainable AI (XAI) has emerged as a field of research with practical and ethical benefits across various domains. This paper not only highlights the advancements in XAI and its application in real-world scenarios but also addresses the ongoing challenges within XAI, emphasizing the need for broader perspectives and collaborative efforts. We bring together experts from diverse fields to identify open problems, striving to synchronize research agendas and accelerate XAI in practical applications. By fostering collaborative discussion and interdisciplinary cooperation, we aim to propel XAI forward, contributing to its continued success. Our goal is to put forward a comprehensive proposal for advancing XAI. To achieve this goal, we present a manifesto of 27 open problems categorized into nine categories. These challenges encapsulate the complexities and nuances of XAI and offer a road map for future research. For each problem, we provide promising research directions in the hope of harnessing the collective intelligence of interested stakeholders.
Interpreting the inner function of neural networks is crucial for the trustworthy development and deployment of these black-box models. Prior interpretability methods focus on correlation-based measures to attribute model decisions to individual examples. However, these measures are susceptible to noise and spurious correlations encoded in the model during the training phase (e.g., biased inputs, model overfitting, or misspecification). Moreover, this process has proven to result in noisy and unstable attributions that prevent any transparent understanding of the model's behavior. In this paper, we develop a robust interventional-based method grounded by causal analysis to capture cause-effect mechanisms in pre-trained neural networks and their relation to the prediction. Our novel approach relies on path interventions to infer the causal mechanisms within hidden layers and isolate relevant and necessary information (to model prediction), avoiding noisy ones. The result is task-specific causal explanatory graphs that can audit model behavior and express the actual causes underlying its performance. We apply our method to vision models trained on classification tasks. On image classification tasks, we provide extensive quantitative experiments to show that our approach can capture more stable and faithful explanations than standard attribution-based methods. Furthermore, the underlying causal graphs reveal the neural interactions in the model, making it a valuable tool in other applications (e.g., model repair).
We introduce a simple non-linear embedding adaptation layer, which is fine-tuned on top of fixed pre-trained features for one-shot tasks, improving significantly transductive entropy-based inference for low-shot regimes. Our norm-induced transformation could be understood as a re-parametrization of the feature space to disentangle the representations of different classes in a task specific manner. It focuses on the relevant feature dimensions while hindering the effects of non-relevant dimensions that may cause overfitting in a one-shot setting. We also provide an interpretation of our proposed feature transformation in the basic case of few-shot inference with K-means clustering. Furthermore, we give an interesting bound-optimization link between K-means and entropy minimization. This emphasizes why our feature transformation is useful in the context of entropy minimization. We report comprehensive experiments, which show consistent improvements over a variety of one-shot benchmarks, outperforming recent state-of-the-art methods.
We introduce a family of interpretable machine learning models, with two broad additions: Linearised Additive Models (LAMs) which replace the ubiquitous logistic link function in General Additive Models (GAMs); and SubscaleHedge, an expert advice algorithm for combining base models trained on subsets of features called subscales. LAMs can augment any additive binary classification model equipped with a sigmoid link function. Moreover, they afford direct global and local attributions of additive components to the model output in probability space. We argue that LAMs and SubscaleHedge improve the interpretability of their base algorithms. Using rigorous null-hypothesis significance testing on a broad suite of financial modelling data, we show that our algorithms do not suffer from large performance penalties in terms of ROC-AUC and calibration.