United States

Explaining the predictions of AI models is paramount in safety-critical applications, such as in legal or medical domains.
One form of explanation for a prediction is an extractive rationale, i.e., a subset of features of an instance that lead the model to give its prediction on the instance. Previous works on generating extractive rationales usually employ a two-phase model: a selector that selects the most important features (i.e., the rationale) followed by a predictor that makes the prediction based exclusively on the selected features. One disadvantage of these works is that the main signal for learning to select features comes from the comparison of the final answers given by the predictor and the ground-truth answers. In this work, we propose to squeeze more information from the predictor via an information calibration method. More precisely, we train two models jointly: one is a typical neural model that solves the task at hand in an accurate but black-box manner, and the other is a selector-predictor model that additionally produces a rationale for its prediction. The first model is used as a guide to the second model. We use an adversarial-based technique to calibrate the information extracted by the two models such that the difference between them is an indicator of the missed or over-selected features. In addition, for natural language tasks, we propose to use a language-model-based regularizer to encourage the extraction of fluent rationales. Experimental results on a sentiment analysis task as well as on three tasks from the legal domain show the effectiveness of our approach to rationale extraction.

AAAI 2021

Learning from the Best: Rationalizing Predictions by Adversarial information Calibration

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

ACL 2024

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

technical paper

The purpose of the AAAI conference is to promote research in artificial intelligence (AI) and scientific exchange among AI researchers, practitioners, scientists, and engineers in affiliated disciplines. 

Most weakly supervised fine-grained image recognition (WFGIR) approaches predominantly focus on learning the discriminative details which contain the visual variances and position clues. The position clues can be indirectly learnt by utilizing context information of discriminative visual content. However, this will cause the selected discriminative regions containing some non-discriminative information introduced by the position clues. These analysis motivate us to directly introduce position clues into visual content to only focus on the visual variances, achieving more precise discriminative region localization. Though important, position modelling usually requires significant pixel/region annotations and therefore is labor-intensive. To address this issue, we propose an end-to-end Dynamic Position-aware Network (DP-Net) to directly incorporate the position clues into visual content and dynamically align them without extra annotations, which eliminates the effect of position information for visual variances of subcategories. In particular, the DP-Net consists of: 1) Position Encoding Module, which learns a set of position-aware parts by directly adding the learnable position information into the horizontal/vertical visual content of images; 2) Position-vision Aligning Module, which dynamically aligns both visual content and learnable position information via performing graph convolution on position-aware parts; 3) Position-vision Reorganization Module, which projects the aligned position clues and visual content into the Euclidean space to construct a position-aware feature maps. Finally, the position-aware feature maps are used which is implicitly applied the aligned visual content and position clues for more accurate discriminative regions localization. Extensive experiments verify that DP-Net yields the best performance under the same settings with most competitive approaches, on CUB Bird, Stanford-Cars, and FGVC Aircraft datasets.

Dynamic Position-Aware Network for Fine-Grained Image Recognition

In this paper, we focus on the challenging few-shot class incremental learning (FSCIL) problem, which requires to transfer knowledge from old tasks to new ones and solves catastrophic forgetting. We propose the exemplar relation distillation incremental learning framework to balance the tasks of old-knowledge preserving and new-knowledge adaptation. First, we construct an exemplar relation graph to represent the knowledge learned by the original network and update gradually for new tasks learning. Then an exemplar relation loss function for discovering the relation knowledge between
different classes is introduced to learn and transfer the structural information in relation graph. A large number of experiments demonstrate that relation knowledge does exist in the exemplars and our approach outperforms other state-of-the-art class-incremental learning methods on the CIFAR100, miniImageNet, and CUB200 datasets.

Few-Shot Class-Incremental Learning via Relation Knowledge Distillation

Controlling the style  of natural language by disentangling the latent space is an important step towards interpretable machine learning. After the latent space is disentangled, the style of a sentence can be transformed by tuning the style representation without affecting other features of the sentence. Previous works usually use adversarial training to guarantee that disentangled vectors do not affect each other.  However, adversarial methods are difficult to train. Especially when there are multiple features (e.g., sentiment, or tense, which we call style types in this paper), each feature requires a separate discriminator  for extracting a disentangled style vector corresponding to that feature. In this paper, we propose a unified distribution-controlling method, which provides each specific style value (the value of style types, e.g., positive sentiment, or past tense) with a unique representation. This method contributes a solid theoretical basis to avoid adversarial training in multi-type disentanglement. We also propose multiple loss functions to achieve a style-content disentanglement as well as a disentanglement among multiple style types. In addition, we observe that if two different style types always have some specific style values    that occur together in the dataset, they will affect each other when transferring the style values. We call this phenomenon  training bias , and we propose a loss function to alleviate such training bias while disentangling multiple types. We conduct experiments on two datasets (Yelp service reviews and Amazon product reviews) to evaluate the style-disentangling effect and the unsupervised style-transfer performance on two style types: sentiment and tense. The experimental results show the effectiveness of our model.

Multi-Type Disentanglement without Adversarial Training

This paper studies the budget-feasible mechanism design over graphs, where a buyer wishes to procure items from sellers, and all participants (the buyer and sellers) can only directly interact with their neighbors during the auction campaign. The problem for the buyer is to use the limited budget to incentivize sellers to propagate auction information to their neighbors, thereby more sellers will be informed of the auction and more item value will be procured. An impossibility result shows that the large-market assumption is necessary. We propose efficient budget-feasible diffusion mechanisms for large markets that simultaneously guarantee individual rationality, budget-feasibility, strong budget-balance, incentive-compatibility to report private costs and diffuse auction information. Moreover, the proposed mechanisms achieve logarithmic approximation that the total procured value is within a logarithmic factor of the optimal solution. Compared to most related budget-feasible mechanisms, which do not take the individual interactions among sellers into account, our mechanisms can incentivize sellers to further propagate auction information to other potential sellers. Meanwhile, existing related diffusion mechanisms only focus on seller-centric auctions and fail to satisfy the budget-feasibility of the buyer.

Budget Feasible Mechanisms over Graphs

Distributed parallel stochastic gradient descent algorithms are workhorses for large scale machine learning tasks. Among them, local stochastic gradient descent (Local SGD) has attracted significant attention due to its low communication complexity. Previous studies prove that the communication complexity of Local SGD with a fixed or an adaptive communication period  is in the order of O (N3/2 T1/2) and O (N3/4 T3/4) when the data distributions on clients are identical (IID) or otherwise (Non-IID), where N is the number of clients and T is the number of iterations. In this paper, to accelerate the convergence by reducing the communication complexity, we propose STagewise Local SGD (STL-SGD), which increases the communication period gradually along with decreasing learning rate. We prove that STL-SGD can keep the same convergence rate and linear speedup as mini-batch SGD. In addition, as the benefit of increasing the communication period, when the objective is strongly convex or satisfies the Polyak-Lojasiewicz condition, the communication complexity of STL-SGD is O (N log T ) and O (N1/2 T1/2) for the IID case and the Non-IID case respectively, achieving significant improvements over Local SGD. Experiments on both convex and non-convex problems demonstrate the superior performance of STL-SGD.

STL-SGD: Speeding Up Local SGD with Stagewise Communication Period

Enabling autonomous systems to comply with an ethical theory is critical given their accelerating deployment in domains that impact society. While many ethical theories have been studied extensively in moral philosophy, they are still challenging to implement by developers who build autonomous systems. This paper proposes a novel approach for building ethically compliant autonomous systems that optimize completing a task while following an ethical framework. First, we introduce a definition of an ethically compliant autonomous system and its properties. Next, we offer a range of ethical frameworks for divine command theory, prima facie duties, and virtue ethics. Finally, we demonstrate the accuracy and usability of our approach in a set of autonomous driving simulations and a user study of planning and robotics experts.

Ethically Compliant Sequential Decision Making

The Right to be Forgotten is part of the recently enacted General Data Protection Regulation (GDPR) law that affects any data holder that has data on European Union residents. It gives EU residents the ability to request deletion of their personal data, including training records used to train machine learning models. Unfortunately, Deep Neural Network models are vulnerable to information leaking attacks such as model inversion attacks which extract class information from a trained model and membership inference attacks which determine the presence of an example in a model's training data. If a malicious party can mount an attack and learn private information that was meant to be removed, then it implies that the model owner has not properly protected their user's rights and their models may not be compliant with the GDPR law. In this paper, we present two efficient methods that address this question of how a model owner or data holder may delete personal data from models in such a way that they may not be vulnerable to model inversion and membership inference attacks while maintaining model efficacy. We start by presenting a real-world threat model that shows that simply removing training data is insufficient to protect users. We follow that up with two data removal methods, namely Unlearning and Amnesiac Unlearning, that enable model owners to protect themselves against such attacks while being compliant with regulations. We provide extensive empirical analysis that show that these methods are indeed efficient, safe to apply, effectively remove learned information about sensitive data from trained models while maintaining model efficacy.

Amnesiac Machine Learning

Motivated by the Bike Angels Program in New York's Citi Bike and Boston's Blue Bikes, we study the use of (registered) volunteers to re-position empty bikes for riders in a bike sharing system. We propose a method that can be used to deploy the volunteers in the system, based on the real time distribution of the bikes in different stations. To account for (random) route demand in the network, we solve a related transshipment network design model and construct a sparse structure to restrict the re-balancing activities of the volunteers (concentrating re-balancing activities on essential routes). We also develop a comprehensive simulation model using a threshold-based policy to deploy the volunteers in real time,  to test the effect of choice restriction on volunteers (suitably deployed) to re-position bikes.  We use the Hubway system in Boston (with 60 stations) to demonstrate that using a sparse structure to concentrate the re-balancing activities of the volunteers, instead of allowing all admissible flows in the system (as in current practice),  can reduce the number of re-balancing moves by a huge amount, losing only a small proportion of demand satisfied.

Bike-Repositioning Using Volunteers: Crowd Sourcing with Choice Restriction

We study the problem of object detection when training and test objects are disjoint, i.e. no training examples of the target classes are available. Existing unseen object detection approaches usually combine generic detection frameworks with a single-path unseen classifier, by aligning object regions with semantic class embeddings. In this paper, inspired from human cognitive experience, we propose a simple but effective dual-path detection model that further explores associative semantics to supplement the basic visual-semantic knowledge transfer. We use a novel target-centric multiple-association strategy to establish concept associations, to ensure that the predictor generalized to unseen domain can be learned during training. In this way, through a reasonable inference fusion mechanism, those two parallel reasoning paths can strengthen the correlation between seen and unseen objects, thus improving detection performance. Experiments show that our inductive method can significantly boost the performance by 7.42% over inductive models, and even 5.25% over transductive models on MSCOCO dataset.

Inference Fusion with Associative Semantics for Unseen Object Detection

Point cloud semantic segmentation is a crucial task in 3D scene understanding. Existing methods mainly focus on employing a large number of annotated labels for supervised semantic segmentation. Nonetheless, manually labeling such large point clouds for the supervised segmentation task is time-consuming. In order to reduce the number of annotated labels, we propose a semi-supervised semantic point cloud segmentation network, named SSPC-Net, where we train the semantic segmentation network by inferring the labels of unlabeled points from the few annotated 3D points. In our method, we first partition the whole point cloud into superpoints and build superpoint graphs to mine the long-range dependencies in point clouds. Based on the constructed superpoint graph, we then develop a dynamic label propagation method to generate the pseudo labels for the unsupervised superpoints. Particularly, we adopt a superpoint dropout strategy to dynamically select the generated pseudo labels. In order to fully exploit the generated pseudo labels of the unsupervised superpoints, we furthermore propose a coupled attention mechanism for superpoint feature embedding. Finally, we employ the cross-entropy loss to train the semantic segmentation network with the labels of the supervised superpoints and the pseudo labels of the unsupervised superpoints. Experiments on various datasets demonstrate that our semisupervised segmentation method can achieve better performance than the current semi-supervised segmentation method with fewer annotated 3D points.

Downloads

Next from AAAI 2021

Dynamic Position-Aware Network for Fine-Grained Image Recognition

Similar lecture

Plausible Extractive Rationalization through Semi-Supervised Entailment Signal

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2021

Dynamic Position-Aware Network for Fine-Grained Image Recognition

Similar lecture

Plausible Extractive Rationalization through Semi-Supervised Entailment Signal

Downloads