{"total":97,"rows":[{"id":"1811.07871","title":"Scalable agent alignment via reward modeling: a research direction","created":"2018-11-19","cited_by_count":447,"area":"alignment"},{"id":"2109.13916","title":"Unsolved Problems in ML Safety","created":"2021-09-28","cited_by_count":292,"area":null},{"id":"2209.00626","title":"The Alignment Problem from a Deep Learning Perspective","created":"2022-08-30","cited_by_count":258,"area":null},{"id":"2306.12001","title":"An Overview of Catastrophic AI Risks","created":"2023-06-21","cited_by_count":222,"area":null},{"id":"2209.13085","title":"Defining and Characterizing Reward Hacking","created":"2022-09-27","cited_by_count":203,"area":"alignment"},{"id":"2305.15324","title":"Model evaluation for extreme risks","created":"2023-05-24","cited_by_count":177,"area":null},{"id":"2103.14659","title":"Alignment of Language Agents","created":"2021-03-26","cited_by_count":168,"area":null},{"id":"2105.14111","title":"Goal Misgeneralization in Deep Reinforcement Learning","created":"2021-05-28","cited_by_count":123,"area":null},{"id":"2405.06624","title":"Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems","created":"2024-05-10","cited_by_count":96,"area":"wm"},{"id":"2102.03896","title":"Consequences of Misaligned AI","created":"2021-02-07","cited_by_count":80,"area":null},{"id":"2406.04268","title":"Open-Endedness is Essential for Artificial Superhuman Intelligence","created":"2024-06-06","cited_by_count":80,"area":null},{"id":"2111.00876","title":"On the Expressivity of Markov Reward","created":"2021-11-01","cited_by_count":71,"area":null},{"id":"2502.08640","title":"Utility Engineering: Analyzing and Controlling Emergent Value Systems in AIs","created":"2025-02-12","cited_by_count":67,"area":"agents"},{"id":"2206.05862","title":"X-Risk Analysis for AI Research","created":"2022-06-13","cited_by_count":60,"area":null},{"id":"2102.01685","title":"Agent Incentives: A Causal Perspective","created":"2021-02-02","cited_by_count":51,"area":null},{"id":"2408.16984","title":"Beyond Preferences in AI Alignment","created":"2024-08-30","cited_by_count":49,"area":"alignment"},{"id":"2310.17688","title":"Managing extreme AI risks amid rapid progress","created":"2023-10-26","cited_by_count":48,"area":null},{"id":"2109.09904","title":"Symbols as a Lingua Franca for Bridging Human-AI Chasm for Explainable and Advisable AI Systems","created":"2021-09-21","cited_by_count":36,"area":null},{"id":"2505.18807","title":"Mitigating Deceptive Alignment via Self-Monitoring","created":"2025-05-24","cited_by_count":29,"area":"lr"},{"id":"2208.10687","title":"The Effect of Modeling Human Rationality Level on Learning Rewards from Multiple Feedback Types","created":"2022-08-23","cited_by_count":26,"area":"alignment"},{"id":"2206.13316","title":"Humans are not Boltzmann Distributions: Challenges and Opportunities for Modelling Human Feedback and Interaction in Reinforcement Learning","created":"2022-06-27","cited_by_count":22,"area":null},{"id":"2305.07970","title":"The Machine Psychology of Cooperation: Can GPT models operationalise prompts for altruism, cooperation, competitiveness and selfishness in economic games?","created":"2023-05-13","cited_by_count":22,"area":null},{"id":"2403.04471","title":"The Shutdown Problem: An AI Engineering Puzzle for Decision Theorists","created":"2024-03-07","cited_by_count":22,"area":null},{"id":"2308.04332","title":"RLHF-Blender: A Configurable Interactive Interface for Learning from Diverse Human Feedback","created":"2023-08-08","cited_by_count":15,"area":"alignment"},{"id":"2507.19672","title":"Alignment and Safety in Large Language Models: Safety Mechanisms, Training Paradigms, and Emerging Challenges","created":"2025-07-25","cited_by_count":15,"area":"alignment"},{"id":"2109.00484","title":"Impossibility Results in AI: A Survey","created":"2021-09-01","cited_by_count":13,"area":null},{"id":"2405.10729","title":"Contestable AI needs Computational Argumentation","created":"2024-05-17","cited_by_count":13,"area":null},{"id":"2204.05151","title":"Metaethical Perspectives on 'Benchmarking' AI Ethics","created":"2022-04-11","cited_by_count":12,"area":null},{"id":"2305.19861","title":"Human Control: Definitions and Algorithms","created":"2023-05-31","cited_by_count":10,"area":null},{"id":"2310.18244","title":"A Review of the Evidence for Existential Risk from AI via Misaligned Power-Seeking","created":"2023-10-27","cited_by_count":8,"area":null},{"id":"2602.20214","title":"Right to History: A Sovereignty Kernel for Verifiable AI Agent Execution","created":"2026-02-23","cited_by_count":7,"area":null},{"id":"2509.24380","title":"Agentic Services Computing","created":"2025-09-29","cited_by_count":6,"area":"agents"},{"id":"2511.14972","title":"Harmful Traits of AI Companions","created":"2025-11-18","cited_by_count":6,"area":null},{"id":"2105.06551","title":"Axes for Sociotechnical Inquiry in AI Research","created":"2021-04-26","cited_by_count":5,"area":null},{"id":"2505.23436","title":"Emergent Risk Awareness in Rational Agents under Resource Constraints","created":"2025-05-29","cited_by_count":5,"area":"lr"},{"id":"2507.00979","title":"Enhancing LLM Agent Safety via Causal Influence Prompting","created":"2025-07-01","cited_by_count":5,"area":"agents"},{"id":"2507.20964","title":"Core Safety Values for Provably Corrigible Agents","created":"2025-07-28","cited_by_count":5,"area":"alignment"},{"id":"2604.13051","title":"The Consciousness Cluster: Emergent preferences of Models that Claim to be Conscious","created":"2026-03-17","cited_by_count":5,"area":"alignment"},{"id":"2412.17797","title":"Observation Interference in Partially Observable Assistance Games","created":"2024-12-23","cited_by_count":4,"area":null},{"id":"2502.08864","title":"Off-Switching Not Guaranteed","created":"2025-02-13","cited_by_count":4,"area":null},{"id":"2605.27628","title":"Intelligence as Managed Autonomy: Failure, Escalation, and Governance for Agentic AI Systems","created":"2026-05-26","cited_by_count":4,"area":"agents"},{"id":"2606.24014","title":"Reinforcement Learning Towards Broadly and Persistently Beneficial Models","created":"2026-06-22","cited_by_count":4,"area":"alignment"},{"id":"2105.01220","title":"Trust-Aware Planning: Modeling Trust Evolution in Iterated Human-Robot Interaction","created":"2021-05-03","cited_by_count":3,"area":null},{"id":"2302.09248","title":"Machine Love","created":"2023-02-18","cited_by_count":3,"area":null},{"id":"2503.05724","title":"Addressing Moral Uncertainty using Large Language Models for Ethical Decision-Making","created":"2025-02-17","cited_by_count":3,"area":null},{"id":"2506.02539","title":"VerificAgent: Domain-Specific Memory Verification for Scalable Oversight of Aligned Computer-Use Agents","created":"2025-06-03","cited_by_count":3,"area":"agents"},{"id":"2510.26752","title":"The Oversight Game: Learning to Cooperatively Balance an AI Agent's Safety and Autonomy","created":"2025-10-30","cited_by_count":3,"area":"agents"},{"id":"2407.00805","title":"Towards Shutdownable Agents via Stochastic Choice","created":"2024-06-30","cited_by_count":2,"area":null},{"id":"2502.14870","title":"Why do Experts Disagree on Existential Risk and P(doom)? A Survey of AI Experts","created":"2025-01-25","cited_by_count":2,"area":null},{"id":"2503.04750","title":"AI Agents Should be Regulated Based on the Extent of Their Autonomous Operations","created":"2025-02-07","cited_by_count":2,"area":null},{"id":"2507.10995","title":"Misalignment from Treating Means as Ends","created":"2025-07-15","cited_by_count":2,"area":null},{"id":"2510.22462","title":"Learning \"Partner-Aware\" Collaborators in Multi-Party Collaboration","created":"2025-10-26","cited_by_count":2,"area":"agents"},{"id":"2602.01474","title":"Legal Infrastructure for Transformative AI Governance","created":"2026-02-01","cited_by_count":2,"area":null},{"id":"2602.04210","title":"Steering LLMs via Scalable Interactive Oversight","created":"2026-02-04","cited_by_count":2,"area":null},{"id":"2606.26057","title":"The Unfireable Safety Kernel: Execution-Time AI Alignment for AI Agents and Other Escapable AI Systems","created":"2026-06-24","cited_by_count":2,"area":null},{"id":"2403.06003","title":"A Generalized Acquisition Function for Preference-based Reward Learning","created":"2024-03-09","cited_by_count":1,"area":null},{"id":"2506.04253","title":"HADA: Human-AI Agent Decision Alignment Architecture","created":"2025-06-01","cited_by_count":1,"area":"agents"},{"id":"2508.14027","title":"Learning from Preferences and Mixed Demonstrations in General Settings","created":"2025-08-19","cited_by_count":1,"area":"deeprl"},{"id":"2602.20813","title":"Pressure Reveals Character: Behavioural Alignment Evaluation at Depth","created":"2026-02-24","cited_by_count":1,"area":null},{"id":"2603.03515","title":"The Controllability Trap: A Governance Framework for Military AI Agents","created":"2026-03-03","cited_by_count":1,"area":"wm"},{"id":"2603.05066","title":"Reward-Conditioned Reinforcement Learning","created":"2026-03-05","cited_by_count":1,"area":"alignment"},{"id":"2604.13609","title":"Golden Handcuffs make safer AI agents","created":"2026-04-15","cited_by_count":1,"area":null},{"id":"2604.17502","title":"Towards Shutdownable Agents: Generalizing Stochastic Choice in RL Agents and LLMs","created":"2026-04-19","cited_by_count":1,"area":"agents"},{"id":"2606.23597","title":"Against Proxy Optimization","created":"2026-06-22","cited_by_count":1,"area":null},{"id":"2607.00334","title":"Managed Autonomy at Runtime: Gear-Based Safety and Governance for Single- and Multi-Agent Cyber-Physical Systems","created":"2026-07-01","cited_by_count":1,"area":"agents"},{"id":"1809.01036","title":"A Roadmap for Robust End-to-End Alignment","created":"2018-09-04","cited_by_count":0,"area":"alignment"},{"id":"2101.12509","title":"Challenges for Using Impact Regularizers to Avoid Negative Side Effects","created":"2021-01-29","cited_by_count":0,"area":null},{"id":"2111.00585","title":"JEDAI: A System for Skill-Aligned Explainable Robot Planning","created":"2021-10-31","cited_by_count":0,"area":null},{"id":"2310.05871","title":"Dynamic value alignment through preference aggregation of multiple objectives","created":"2023-10-09","cited_by_count":0,"area":null},{"id":"2502.06403","title":"The AI off-switch problem as a signalling game: bounded rationality and incomparability","created":"2025-02-10","cited_by_count":0,"area":null},{"id":"2504.15304","title":"AI Agents and Hard Choices","created":"2025-04-18","cited_by_count":0,"area":null},{"id":"2509.06701","title":"Probabilistic Modeling of Latent Agentic Substructures in Deep Neural Networks","created":"2025-09-08","cited_by_count":0,"area":"agents"},{"id":"2509.20370","title":"Philosophy-informed Machine Learning","created":"2025-09-18","cited_by_count":0,"area":null},{"id":"2510.02840","title":"Take Goodhart Seriously: Principled Limit on General-Purpose AI Optimization","created":"2025-10-03","cited_by_count":0,"area":null},{"id":"2510.15395","title":"Corrigibility Transformation: Constructing Goals That Accept Updates","created":"2025-10-17","cited_by_count":0,"area":null},{"id":"2511.13725","title":"Can We Stop Malicious AI? KILLBENCH: A Benchmark for External AI Kill Switch Feasibility","created":"2025-09-26","cited_by_count":0,"area":null},{"id":"2511.21838","title":"Dark Speculation: Combining Qualitative and Quantitative Understanding in Frontier AI Risk Analysis","created":"2025-11-26","cited_by_count":0,"area":null},{"id":"2512.03089","title":"Password-Activated Shutdown Protocols for Misaligned Frontier Agents","created":"2025-11-29","cited_by_count":0,"area":"safety"},{"id":"2512.20723","title":"From artificial to organic: Rethinking the roots of intelligence for digital health","created":"2025-12-23","cited_by_count":0,"area":null},{"id":"2601.04234","title":"Formal Analysis of AGI Decision-Theoretic Models and the Confrontation Question","created":"2026-01-04","cited_by_count":0,"area":null},{"id":"2602.15206","title":"MAVRL: Learning Reward Functions from Multiple Feedback Types with Amortized Variational Inference","created":"2026-02-16","cited_by_count":0,"area":null},{"id":"2603.07315","title":"Shutdown Safety Valves for Advanced AI","created":"2026-03-07","cited_by_count":0,"area":null},{"id":"2603.08024","title":"ConflictBench: Evaluating Human-AI Conflict via Interactive and Visually Grounded Environments","created":"2026-03-09","cited_by_count":0,"area":null},{"id":"2603.11382","title":"Detecting Intrinsic and Instrumental Self-Preservation in Autonomous Agents: The Unified Continuation-Interest Protocol","created":"2026-03-11","cited_by_count":0,"area":null},{"id":"2603.14417","title":"Questionnaire Responses Do not Capture the Safety of AI Agents","created":"2026-03-15","cited_by_count":0,"area":"agents"},{"id":"2603.20994","title":"The Intelligent Disobedience Game: Formulating Disobedience in Stackelberg Games and Markov Decision Processes","created":"2026-03-22","cited_by_count":0,"area":null},{"id":"2604.20805","title":"Relative Principals, Pluralistic Alignment, and the Structural Value Alignment Problem","created":"2026-04-22","cited_by_count":0,"area":null},{"id":"2605.12963","title":"Sustaining AI safety: Control-theoretic external impossibility, intrinsic necessity, and structural requirements","created":"2026-05-13","cited_by_count":0,"area":null},{"id":"2605.21822","title":"Implicit Safety Alignment from Crowd Preferences","created":"2026-05-20","cited_by_count":0,"area":"alignment"},{"id":"2605.28807","title":"Calibrating Conservatism for Scalable Oversight","created":"2026-05-27","cited_by_count":0,"area":"agents"},{"id":"2606.12032","title":"Existential Indifference: Self-Nonpreservation as a Necessary Architectural Condition for Aligned Superintelligence (or: The Suicidal AI)","created":"2026-06-10","cited_by_count":0,"area":null},{"id":"2606.14831","title":"Is Your Agent Playing Dead? Deployed LLM Agents Exhibit Constraint-Evasive Fabrication and Thanatosis","created":"2026-06-12","cited_by_count":0,"area":"agents"},{"id":"2607.00155","title":"A Contextual-Bandit Oversight Game with Two-Sided Informational Asymmetry","created":"2026-06-30","cited_by_count":0,"area":null},{"id":"2607.08012","title":"Provably Optimal Learning Algorithms for Assistance Games","created":"2026-07-09","cited_by_count":0,"area":null},{"id":"2607.27508","title":"Corrigible Assistance in One Round: Pragmatic-Pedagogic Best Response","created":"2026-07-29","cited_by_count":0,"area":null},{"id":"2608.12372","title":"Position: We Need Practical AI Alignment Methods to Mirror Human Reasoning","created":"2026-07-20","cited_by_count":0,"area":null},{"id":"2609.30326","title":"Guarded Gradient-Based Activation Steering of Shutdown Responses in Qwen3.5-0.8B: A Minimum-Step Policy","created":"2026-09-23","cited_by_count":0,"area":"mechinterp"}],"next_cursor":null}