AAAI-2024-Papers

Application

Safe, Robust and Responsible AI

:id:	Title	Repo	Paper	Video
	ImageCaptioner2: Image Captioner for Image Captioning Bias Amplification Assessment	:heavy_minus_sign:
	A Framework for Data-Driven Explainability in Mathematical Optimization	:heavy_minus_sign:
	On the Importance of Application-Grounded Experimental Design for Evaluating Explainable ML Methods	:heavy_minus_sign:
	Risk-Aware Continuous Control with Neural Contextual Bandits	:heavy_minus_sign:
	Robust Uncertainty Quantification Using Conformalised Monte Carlo Prediction	:heavy_minus_sign:
	CCTR: Calibrating Trajectory Prediction for Uncertainty-Aware Motion Planning in Autonomous Driving	:heavy_minus_sign:
	Rethinking the Development of Large Language Models from the Causal Perspective: A Legal Text Prediction Case Study	:heavy_minus_sign:
	Truth Forest: Toward Multi-Scale Truthfulness in Large Language Models through Intervention without Tuning	:heavy_minus_sign:
	Constrained Meta-Reinforcement Learning for Adaptable Safety Guarantee with Differentiable Convex Programming	:heavy_minus_sign:
	Conformal Prediction Regions for Time Series Using Linear Complementarity Programming	:heavy_minus_sign:
	TTTS: Tree Test Time Simulation for Enhancing Decision Tree Robustness against Adversarial Examples	:heavy_minus_sign:
	Find the Lady: Permutation and Re-synchronization of Deep Neural Networks	:heavy_minus_sign:
	Stability Analysis of Switched Linear Systems with Neural Lyapunov Functions	:heavy_minus_sign:
	Robustness Verification of Multi-Class Tree Ensembles	:heavy_minus_sign:
	P2BPO: Permeable Penalty Barrier-Based Policy Optimization for Safe RL	:heavy_minus_sign:
	Trade-Offs in Fine-Tuned Diffusion Models between Accuracy and Interpretability	:heavy_minus_sign:
	From Hope to Safety: Unlearning Biases of Deep Models via Gradient Penalization in Latent Space	:heavy_minus_sign:
	Automatically Testing Functional Properties of Code Translation Models	:heavy_minus_sign:
	A Simple and Yet Fairly Effective Defense for Graph Neural Networks	:heavy_minus_sign:
	Invisible Backdoor Attack against 3D Point Cloud Classifier in Graph Spectral Domain	:heavy_minus_sign:
	CASE: Exploiting Intra-class Compactness and Inter-class Separability of Feature Embeddings for Out-of-Distribution Detection	:heavy_minus_sign:
	Solving Non-rectangular Reward-Robust MDPs via Frequency Regularization	:heavy_minus_sign:
	Balance Reward and Safety Optimization for Safe Reinforcement Learning: A Perspective of Gradient Manipulation	:heavy_minus_sign:
	π-Light: Programmatic Interpretable Reinforcement Learning for Resource-Limited Traffic Signal Control	:heavy_minus_sign:
	Generative Model for Decision Trees	:heavy_minus_sign:
	Omega-Regular Decision Processes	:heavy_minus_sign:
	Provable Robustness against a Union of L_0 Adversarial Attacks	:heavy_minus_sign:
	All but One: Surgical Concept Erasing with Model Preservation in Text-to-Image Diffusion Models	:heavy_minus_sign:
	Towards Efficient Verification of Quantized Neural Networks	:heavy_minus_sign:
	On the Concept Trustworthiness in Concept Bottleneck Models	:heavy_minus_sign:
	Personalization as a Shortcut for Few-Shot Backdoor Attack against Text-to-Image Diffusion Models	:heavy_minus_sign:
	Stronger and Transferable Node Injection Attacks	:heavy_minus_sign:
	Learning Fair Policies for Multi-Stage Selection Problems from Observational Data	:heavy_minus_sign:
	NeRFail: Neural Radiance Fields-Based Multiview Adversarial Attack	:heavy_minus_sign:
	Analysis of Differentially Private Synthetic Data: A Measurement Error Approach	:heavy_minus_sign:
	Chasing Fairness in Graphs: A GNN Architecture Perspective	:heavy_minus_sign:
	Assume-Guarantee Reinforcement Learning	:heavy_minus_sign:
	DeepBern-Nets: Taming the Complexity of Certifying Neural Networks Using Bernstein Polynomial Activations and Precise Bound Propagation	:heavy_minus_sign:
	Layer Attack Unlearning: Fast and Accurate Machine Unlearning via Layer Level Attack and Knowledge Distillation	:heavy_minus_sign:
	Quilt: Robust Data Segment Selection against Concept Drifts	:heavy_minus_sign:
	OUTFOX: LLM-Generated Essay Detection Through In-Context Learning with Adversarially Generated Examples	:heavy_minus_sign:
	Accelerating Adversarially Robust Model Selection for Deep Neural Networks via Racing	:heavy_minus_sign:
	Robust Active Measuring under Model Uncertainty	:heavy_minus_sign:
	Towards Large Certified Radius in Randomized Smoothing Using Quasiconcave Optimization	:heavy_minus_sign:
	Contrastive Credibility Propagation for Reliable Semi-supervised Learning	:heavy_minus_sign:
	Exponent Relaxation of Polynomial Zonotopes and Its Applications in Formal Neural Network Verification	:heavy_minus_sign:
	I Prefer Not to Say: Protecting User Consent in Models with Optional Personal Data	:heavy_minus_sign:
	Promoting Counterfactual Robustness through Diversity	:heavy_minus_sign:
	Revisiting the Information Capacity of Neural Network Watermarks: Upper Bound Estimation and Beyond	:heavy_minus_sign:
	PointCVaR: Risk-Optimized Outlier Removal for Robust 3D Point Cloud Classification	:heavy_minus_sign:
	Game-Theoretic Unlearnable Example Generator	:heavy_minus_sign:
	Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning	:heavy_minus_sign:
	Handling Long and Richly Constrained Tasks through Constrained Hierarchical Reinforcement Learning	:heavy_minus_sign:
	Combining Graph Transformers Based Multi-Label Active Learning and Informative Data Augmentation for Chest Xray Classification	:heavy_minus_sign:
	Enumerating Safe Regions in Deep Neural Networks with Provable Probabilistic Guarantees	:heavy_minus_sign:
	Divide-and-Aggregate Learning for Evaluating Performance on Unlabeled Data	:heavy_minus_sign:
	SentinelLMs: Encrypted Input Adaptation and Fine-Tuning of Language Models for Private and Secure Inference	:heavy_minus_sign:
	Safeguarded Progress in Reinforcement Learning: Safe Bayesian Exploration for Control Policy Synthesis	:heavy_minus_sign:
	Feature Unlearning for Pre-trained GANs and VAEs	:heavy_minus_sign:
	Reward Certification for Policy Smoothed Reinforcement Learning	:heavy_minus_sign:
	EncryIP: A Practical Encryption-Based Framework for Model Intellectual Property Protection	:heavy_minus_sign:
	Neural Closure Certificates	:heavy_minus_sign:
	SocialStigmaQA: A Benchmark to Uncover Stigma Amplification in Generative Language Models	:heavy_minus_sign:
	MaxEnt Loss: Constrained Maximum Entropy for Calibration under Out-of-Distribution Shift	:heavy_minus_sign:
	ORES: Open-Vocabulary Responsible Visual Synthesis	:heavy_minus_sign:
	Q-SENN: Quantized Self-Explaining Neural Networks	:heavy_minus_sign:
	Understanding Likelihood of Normalizing Flow and Image Complexity through the Lens of Out-of-Distribution Detection	:heavy_minus_sign:
	Adversarial Initialization with Universal Adversarial Perturbation: A New Approach to Fast Adversarial Training	:heavy_minus_sign:
	A PAC Learning Algorithm for LTL and Omega-Regular Objectives in MDPs	:heavy_minus_sign:
	Robust Stochastic Graph Generator for Counterfactual Explanations	:heavy_minus_sign:
	Visual Adversarial Examples Jailbreak Aligned Large Language Models	:heavy_minus_sign:
	Dissenting Explanations: Leveraging Disagreement to Reduce Model Overreliance	:heavy_minus_sign:
	I-CEE: Tailoring Explanations of Image Classification Models to User Expertise	:heavy_minus_sign:
	A Simple and Practical Method for Reducing the Disparate Impact of Differential Privacy	:heavy_minus_sign:
	Interpretability Benchmark for Evaluating Spatial Misalignment of Prototypical Parts Explanations	:heavy_minus_sign:
	Human-Guided Moral Decision Making in Text-Based Games	:heavy_minus_sign:
	Towards Fairer Centroids in K-means Clustering	:heavy_minus_sign:
	Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise	:heavy_minus_sign:
	Bidirectional Contrastive Split Learning for Visual Question Answering	:heavy_minus_sign:
	Quantile-Based Maximum Likelihood Training for Outlier Detection	:heavy_minus_sign:
	Sparsity-Guided Holistic Explanation for LLMs with Interpretable Inference-Time Intervention	:heavy_minus_sign:
	Toward More Generalized Malicious URL Detection Models	:heavy_minus_sign:
	Self-Supervised Likelihood Estimation with Energy Guidance for Anomaly Segmentation in Urban Scenes	:heavy_minus_sign:
	Pure-Past Action Masking	:heavy_minus_sign:
	Long-Term Safe Reinforcement Learning with Binary Feedback	:heavy_minus_sign:
	Identifying Reasons for Bias: An Argumentation-Based Approach	:heavy_minus_sign:
	Would You Like Your Data to Be Trained? A User Controllable Recommendation Framework	:heavy_minus_sign:
	Moderate Message Passing Improves Calibration: A Universal Way to Mitigate Confidence Bias in Graph Neural Networks	:heavy_minus_sign:
	Generating Diagnostic and Actionable Explanations for Fair Graph Neural Networks	:heavy_minus_sign:
	Physics-Informed Representation and Learning: Control and Risk Quantification	:heavy_minus_sign:
	Safe Reinforcement Learning with Instantaneous Constraints: The Role of Aggressive Exploration	:heavy_minus_sign:
	Concealing Sensitive Samples against Gradient Leakage in Federated Learning	:heavy_minus_sign:
	The Evidence Contraction Issue in Deep Evidential Regression: Discussion and Solution	:heavy_minus_sign:
	Byzantine-Robust Decentralized Learning via Remove-then-Clip Aggregation	:heavy_minus_sign:
	Hypothesis Testing for Class-Conditional Noise Using Local Maximum Likelihood	:heavy_minus_sign:
	Providing Fair Recourse over Plausible Groups	:heavy_minus_sign:
	Representation-Based Robustness in Goal-Conditioned Reinforcement Learning	:heavy_minus_sign:
	Enhancing Off-Policy Constrained Reinforcement Learning through Adaptive Ensemble C Estimation	:heavy_minus_sign:
	Efficient Toxic Content Detection by Bootstrapping and Distilling Large Language Models	:heavy_minus_sign:
	LR-XFL: Logical Reasoning-Based Explainable Federated Learning	:heavy_minus_sign:
	GaLileo: General Linear Relaxation Framework for Tightening Robustness Certification of Transformers	:heavy_minus_sign:
	A Huber Loss Minimization Approach to Byzantine Robust Federated Learning	:heavy_minus_sign:
	Responsible Bandit Learning via Privacy-Protected Mean-Volatility Utility	:heavy_minus_sign:
	UMA: Facilitating Backdoor Scanning via Unlearning-Based Model Ablation	:heavy_minus_sign:
	AdvST: Revisiting Data Augmentations for Single Domain Generalization	:heavy_minus_sign:
	Can LLM Replace Stack Overflow? A Study on Robustness and Reliability of Large Language Model Code Generation	:heavy_minus_sign:
	DataElixir: Purifying Poisoned Dataset to Mitigate Backdoor Attacks via Diffusion Models	:heavy_minus_sign:
	Closing the Gap: Achieving Better Accuracy-Robustness Tradeoffs against Query-Based Attacks	:heavy_minus_sign:
	Coevolutionary Algorithm for Building Robust Decision Trees under Minimax Regret	:heavy_minus_sign: