AI Safety

ToolSword: Unveiling Safety Issues of Large Language Models in Tool Learning Across Three Stages ACL 2024 Towards a Unified Framework for Adaptable Problematic Content Detection via Continual Learning NAACL 2024 Towards Faithful XAI Evaluation via Generalization-Limited Backdoor Watermark ICLR 2024 Towards General Robustness Verification of MaxPool-based Convolutional Neural Networks via Tightening Linear Approximation CVPR 2024 Towards Healthy AI: Large Language Models Need Therapists Too NAACL 2024 Towards Implicit Bias Detection and Mitigation in Multi-Agent LLM Interactions EMNLP 2024 Towards Probing Speech-Specific Risks in Large Multimodal Models: A Taxonomy, Benchmark, and Insights EMNLP 2024 Towards Safe Concept Transfer of Multi-Modal Diffusion via Causal Representation Editing NeurIPS 2024 Towards safe multi-task Bayesian optimization L4DC 2024 Towards Safer Large Language Models through Machine Unlearning ACL 2024 Towards the Robustness of Differentially Private Federated Learning AAAI 2024 Towards Tool Use Alignment of Large Language Models EMNLP 2024 Towards Tracing Trustworthiness Dynamics: Revisiting Pre-training Period of Large Language Models ACL 2024 Towards Trustworthy Autonomous Systems via Conversations and Explanations AAAI 2024 Towards Trustworthy Deep Learning AAAI 2024 Towards Understanding Jailbreak Attacks in LLMs: A Representation Space Analysis EMNLP 2024 Towards Understanding Sycophancy in Language Models ICLR 2024 Training Socially Aligned Language Models on Simulated Social Interactions ICLR 2024 Trap-MID: Trapdoor-based Defense against Model Inversion Attacks NeurIPS 2024 Treatment of Statistical Estimation Problems in Randomized Smoothing for Adversarial Robustness NeurIPS 2024 Tree of Attacks: Jailbreaking Black-Box LLMs Automatically NeurIPS 2024 Tricking LLMs into Disobedience: Formalizing, Analyzing, and Detecting Jailbreaks COLING 2024 TrojFSP: Trojan Insertion in Few-shot Prompt Tuning NAACL 2024 Truly No-Regret Learning in Constrained MDPs ICML 2024 TrustAgent: Towards Safe and Trustworthy LLM-based Agents EMNLP 2024