{"list":"other","total":76,"rows":[{"id":"2303.08774","title":"GPT-4 Technical Report","created":"2023-03-15","cited_by_count":18396},{"id":"2203.02155","title":"Training language models to follow instructions with human feedback","created":"2022-03-04","cited_by_count":15971},{"id":"2306.05685","title":"Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena","created":"2023-06-09","cited_by_count":8626},{"id":"1412.6572","title":"Explaining and Harnessing Adversarial Examples","created":"2014-12-20","cited_by_count":7364},{"id":"1706.06083","title":"Towards Deep Learning Models Resistant to Adversarial Attacks","created":"2017-06-19","cited_by_count":6415},{"id":"1312.6199","title":"Intriguing properties of neural networks","created":"2013-12-21","cited_by_count":5155},{"id":"1606.05250","title":"SQuAD: 100,000+ Questions for Machine Comprehension of Text","created":"2016-06-16","cited_by_count":4148},{"id":"1706.03741","title":"Deep reinforcement learning from human preferences","created":"2017-06-12","cited_by_count":3912},{"id":"2204.05862","title":"Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback","created":"2022-04-12","cited_by_count":3720},{"id":"2307.15043","title":"Universal and Transferable Adversarial Attacks on Aligned Language Models","created":"2023-07-27","cited_by_count":3104},{"id":"1608.04644","title":"Towards Evaluating the Robustness of Neural Networks","created":"2016-08-16","cited_by_count":2998},{"id":"2101.00027","title":"The Pile: An 800GB Dataset of Diverse Text for Language Modeling","created":"2020-12-31","cited_by_count":2411},{"id":"2307.02483","title":"Jailbroken: How Does LLM Safety Training Fail?","created":"2023-07-05","cited_by_count":1651},{"id":"2302.12173","title":"Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection","created":"2023-02-23","cited_by_count":1443},{"id":"2402.04249","title":"HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal","created":"2024-02-06","cited_by_count":1429},{"id":"2310.08419","title":"Jailbreaking Black Box Large Language Models in Twenty Queries","created":"2023-10-12","cited_by_count":1420},{"id":"1902.02918","title":"Certified Adversarial Robustness via Randomized Smoothing","created":"2019-02-08","cited_by_count":1340},{"id":"2312.06674","title":"Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations","created":"2023-12-07","cited_by_count":1207},{"id":"2310.03693","title":"Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!","created":"2023-10-05","cited_by_count":1154},{"id":"1912.03817","title":"Machine Unlearning","created":"2019-12-09","cited_by_count":1061},{"id":"2406.11717","title":"Refusal in Language Models Is Mediated by a Single Direction","created":"2024-06-17","cited_by_count":1032},{"id":"1704.03976","title":"Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning","created":"2017-04-13","cited_by_count":920},{"id":"2112.03570","title":"Membership Inference Attacks From First Principles","created":"2021-12-07","cited_by_count":878},{"id":"2307.15217","title":"Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback","created":"2023-07-27","cited_by_count":759},{"id":"2107.06499","title":"Deduplicating Training Data Makes Language Models Better","created":"2021-07-14","cited_by_count":714},{"id":"1708.06131","title":"Evasion Attacks against Machine Learning at Test Time","created":"2017-08-21","cited_by_count":689},{"id":"2306.05499","title":"Prompt Injection attack against LLM-integrated Applications","created":"2023-06-08","cited_by_count":605},{"id":"2401.06373","title":"How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs","created":"2024-01-12","cited_by_count":573},{"id":"2002.08347","title":"On Adaptive Attacks to Adversarial Example Defenses","created":"2020-02-19","cited_by_count":555},{"id":"2404.01318","title":"JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models","created":"2024-03-28","cited_by_count":536},{"id":"2403.03218","title":"The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning","created":"2024-03-05","cited_by_count":509},{"id":"2404.13208","title":"The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions","created":"2024-04-19","cited_by_count":493},{"id":"2404.02151","title":"Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks","created":"2024-04-02","cited_by_count":486},{"id":"1902.06705","title":"On Evaluating Adversarial Robustness","created":"2019-02-18","cited_by_count":447},{"id":"2310.16789","title":"Detecting Pretraining Data from Large Language Models","created":"2023-10-25","cited_by_count":387},{"id":"2402.10260","title":"A StrongREJECT for Empty Jailbreaks","created":"2024-02-15","cited_by_count":385},{"id":"2310.02238","title":"Who's Harry Potter? Approximate Unlearning in LLMs","created":"2023-10-03","cited_by_count":381},{"id":"2306.15447","title":"Are aligned neural networks adversarially aligned?","created":"2023-06-26","cited_by_count":315},{"id":"2305.00944","title":"Poisoning Language Models During Instruction Tuning","created":"2023-05-01","cited_by_count":269},{"id":"2407.06460","title":"MUSE: Machine Unlearning Six-Way Evaluation for Language Models","created":"2024-07-08","cited_by_count":261},{"id":"2405.20947","title":"OR-Bench: An Over-Refusal Benchmark for Large Language Models","created":"2024-05-31","cited_by_count":244},{"id":"2402.05162","title":"Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications","created":"2024-02-07","cited_by_count":231},{"id":"2311.03348","title":"Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation","created":"2023-11-06","cited_by_count":214},{"id":"2404.09932","title":"Foundational Challenges in Assuring Alignment and Safety of Large Language Models","created":"2024-04-15","cited_by_count":199},{"id":"2402.07841","title":"Do Membership Inference Attacks Work on Large Language Models?","created":"2024-02-12","cited_by_count":195},{"id":"2501.18837","title":"Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming","created":"2025-01-31","cited_by_count":181},{"id":"2012.10544","title":"Dataset Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses","created":"2020-12-18","cited_by_count":180},{"id":"2407.15549","title":"Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs","created":"2024-07-22","cited_by_count":173},{"id":"2402.16835","title":"Eight Methods to Evaluate Robust Unlearning in LLMs","created":"2024-02-26","cited_by_count":163},{"id":"2309.17410","title":"Can Sensitive Information Be Deleted From LLMs? Objectives for Defending Against Extraction Attacks","created":"2023-09-29","cited_by_count":158},{"id":"2408.15221","title":"LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yet","created":"2024-08-27","cited_by_count":151},{"id":"2308.12284","title":"D4: Improving LLM Pretraining via Document De-Duplication and Diversification","created":"2023-08-23","cited_by_count":150},{"id":"2406.13352","title":"AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents","created":"2024-06-19","cited_by_count":145},{"id":"2408.00761","title":"Tamper-Resistant Safeguards for Open-Weight LLMs","created":"2024-08-01","cited_by_count":138},{"id":"2411.10414","title":"Llama Guard 3 Vision: Safeguarding Human-AI Image Understanding Conversations","created":"2024-11-15","cited_by_count":129},{"id":"2311.14455","title":"Universal Jailbreak Backdoors from Poisoned Human Feedback","created":"2023-11-24","cited_by_count":128},{"id":"2401.14446","title":"Black-Box Access is Insufficient for Rigorous AI Audits","created":"2024-01-25","cited_by_count":126},{"id":"2402.14016","title":"Is LLM-as-a-Judge Robust? Investigating Universal Adversarial Attacks on Zero-shot LLM Assessment","created":"2024-02-21","cited_by_count":126},{"id":"2409.18025","title":"An Adversarial Perspective on Machine Unlearning for AI Safety","created":"2024-09-26","cited_by_count":123},{"id":"1809.02104","title":"Are adversarial examples inevitable?","created":"2018-09-06","cited_by_count":113},{"id":"2403.05030","title":"Defending Against Unforeseen Failure Modes with Latent Adversarial Training","created":"2024-03-08","cited_by_count":88},{"id":"2403.03792","title":"Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks","created":"2024-03-06","cited_by_count":85},{"id":"2412.03556","title":"Best-of-N Jailbreaking","created":"2024-12-04","cited_by_count":76},{"id":"2406.16201","title":"Blind Baselines Beat Membership Inference Attacks for Foundation Models","created":"2024-06-23","cited_by_count":69},{"id":"2501.18841","title":"Trading Inference-Time Compute for Adversarial Robustness","created":"2025-01-31","cited_by_count":61},{"id":"1807.06732","title":"Motivating the Rules of the Game for Adversarial Example Research","created":"2018-07-18","cited_by_count":57},{"id":"2310.15007","title":"Did the Neurons Read your Book? Document-level Membership Inference for Large Language Models","created":"2023-10-23","cited_by_count":56},{"id":"2402.15911","title":"PRP: Propagating Universal Perturbations to Attack Large Language Model Guard-Rails","created":"2024-02-24","cited_by_count":52},{"id":"2412.07097","title":"On Evaluating the Durability of Safeguards for Open-Weight LLMs","created":"2024-12-10","cited_by_count":52},{"id":"2407.00106","title":"UnUnlearning: Unlearning is not sufficient for content regulation in advanced generative AI","created":"2024-06-27","cited_by_count":51},{"id":"2409.19798","title":"Membership Inference Attacks Cannot Prove that a Model Was Trained On Your Data","created":"2024-09-29","cited_by_count":50},{"id":"2410.13722","title":"Persistent Pre-Training Poisoning of LLMs","created":"2024-10-17","cited_by_count":48},{"id":"2410.12855","title":"JAILJUDGE: A Comprehensive Jailbreak Judge Benchmark with Multi-Agent Enhanced Explanation Evaluation Framework","created":"2024-10-11","cited_by_count":47},{"id":"2406.12027","title":"Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI","created":"2024-06-17","cited_by_count":38},{"id":"2407.02551","title":"Breach By A Thousand Leaks: Unsafe Information Leakage in `Safe' AI Responses","created":"2024-07-02","cited_by_count":18},{"id":"2410.03489","title":"Gradient-based Jailbreak Images for Multimodal Fusion Models","created":"2024-10-04","cited_by_count":14}],"next_cursor":null}