Key papers

• AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security (2026)
  Dongrui Liu et al.
  14.67
• AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration (2026)
  Jiaqi Liu et al.
  13.41
• BadWAM: When World-Action Models Dream Right but Act Wrong (2026)
  Qi Li et al.
  12.78
• The Verification Horizon: No Silver Bullet for Coding Agent Rewards (2026)
  Binghai Wang et al.
  12.31
• AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security (2026)
  Dongrui Liu et al.
  12.01
• Prompt Injection Attacks in Large Language Models and AI Agent Systems: A Comprehensive Review of Vulnerabilities, Attack Vectors, and Defense Mechanisms (2026)
  Saidakhror Gulyamov et al.
  11.81
• It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs (2026)
  Sangwoo Park et al.
  11.53
• Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It (2026)
  Yupu Hao et al.
  10.37
• Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems (2026)
  Jiacheng Liu et al.
  10.29
• Confidence-Aware Tool Orchestration for Robust Video Understanding (2026)
  Yangfan He et al.
  10.05
• From Prompt Injection to Persistent Control: Defending Agentic Harness Against Trojan Backdoors (2026)
  Jiejun Tan et al.
  10.04
• Critique of Agent Model (2026)
  Eric Xing et al.
  9.48
• RedAct: Redacting Agent Capability Traces for Procedural Skill Protection (2026)
  Shuwen Xu et al.
  9.37
• Shieldstral (2026)
  Antonia Calvi et al.
  9.27
• OmniVerifier-M1: Multimodal Meta-Verifier with Explicit Structured Recalibration (2026)
  Xinchen Zhang et al.
  9.17
• SeerGuard: A Safety Framework for Mobile GUI Agents via World Model Prediction (2026)
  Xue Yu et al.
  9.05
• SingGuard: A Policy-Adaptive Multimodal LLM Guardrail with Dynamic Reasoning (2026)
  SingGuard Team
  9.00
• MultAttnAttrib: Training-Free Multimodal Attribution in Long Document Question Answering (2026)
  Dang Quang Thien Tran et al.
  8.81
• GPT-Red: Automated Red Teaming via Self-Play at Scale (2026)
  Eric Wallace et al.
  8.81
• Beyond Final Answers: Auditing Trajectory-Level Hallucinations in Multi-Agent Industrial Workflows (2026)
  Harshada Badave et al.
  8.69
• ORCA-bench: How Ready Are Language Model Agents for Oncall? (2026)
  Albert Gong et al.
  8.66
• Steering LLMs via Scalable Interactive Oversight (2026)
  Enyu Zhou et al.
  8.00
• A Survey on Agent Workflow -- Status and Future (2025)
  Chaojia Yu et al.
  7.99
• Safety Testing LLM Agents at Scale: From Risk Discovery to Evidence-Grounded Verification (2026)
  Yunhao Feng et al.
  7.92
• BAPO: Boundary-Aware Policy Optimization for Reliable Agentic Search (2026)
  Shiyu Liu et al.
  7.82
• Distributed Attacks in Persistent-State AI Control (2026)
  Josh Hills et al.
  7.82
• PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models (2026)
  Ziliang Zhao et al.
  7.68
• Learning to Act under Noise: Enhancing Agent Robustness via Noisy Environments (2026)
  Yuxin Chen et al.
  7.68
• Agent Security is a Systems Problem (2026)
  Mihai Christodorescu et al.
  7.66
• Governing Evolving Memory in LLM Agents: Risks, Mechanisms, and the Stability and Safety Governed Memory (SSGM) Framework (2026)
  Chingkwun Lam et al.
  7.56
• Intelligent AI Delegation (2026)
  Nenad Toma\v{s}ev et al.
  7.50
• StealthBench: Measuring Operational Stealth in Autonomous Offensive-Security Agents (2026)
  Ads Dawson et al.
  7.32
• PolicyGuard: A Dialogue-Grounded Sub-Agent Verifier for Policy Adherence in LLM Agents (2026)
  Seongjae Kang et al.
  7.25
• Emergent Languages in Populations of Language Model Agents: From Token Efficiency to Oversight Evasion (2026)
  Stine Lyngs{\o} Beltoft et al.
  7.20
• Evaluating routing stability and coordination in swarm-based multi-agent task-oriented dialogue systems (2026)
  Abuzar Khan et al.
  7.13
• Tracing Agentic Failure from the Flow of Success (2026)
  Samuel Yeh et al.
  6.89
• Enhancing Robustness of LLM-Driven Multi-Agent Systems through Randomized Smoothing (2025)
  Jinwei Hu et al.
  6.80
• Agent libOS: A Library-OS-Inspired Runtime for Long-Running, Capability-Controlled LLM Agents (2026)
  Yingqi Zhang
  6.64
• The Meta-Agent Challenge: Are Current Agents Capable of Autonomous Agent Development? (2026)
  Xinyu Lu et al.
  6.64
• Agentic Uncertainty Quantification (2026)
  Jiaxin Zhang et al.
  6.56
• Constitutional Midtraining: Content Presence Drives Alignment Gains (2026)
  Desiree Cho et al.
  6.44
• Agora: Toward Autonomous Bug Detection in Production-Level Consensus Protocols with LLM Agents (2026)
  Xiang Liu et al.
  6.35
• Safety Aware Task Planning Via Large Language Models In Robotics (2025)
  Azal Ahmad Khan, Michael Andrev, Muhammad Ali Murtaza, et al.
  6.28
• AI Chatbot Suicide Risk Detection and Response: Human Validation Study of the Open-Source VERA-MH Safety Evaluation (2026)
  Kate H. Bentley et al.
  6.18
• Attack the Messages, Not the Agents: A Multi-round Adaptive Stealthy Tampering Framework for LLM-MAS (2025)
  Bingyu Yan et al.
  6.12
• Prismata: Confining Cross-Site Prompt Injection in Web Agents (2026)
  Corban Villa et al.
  5.84
• !Imperio, smolVLA: The Implications of Data Poisoning on Open Source Robotics (2026)
  Stefan B\"uhler et al.
  5.82
• Validate the Dream Before You Trust Its Verdict: Admissibility for World-Model Simulators (2026)
  Christian Oefinger et al.
  5.82
• AgentAbstain: Do LLM Agents Know When Not to Act? (2026)
  Xun Liu et al.
  5.82
• Coercion and Deception in AI-to-AI Management: An Agentic Benchmark of Unprompted Escalation (2026)
  Jasmine Brazilek et al.
  5.81
• Train the Model, Not the Reader: Decodability Supervision for Verifiable Activation Explanations (2026)
  Hiskias Dingeto
  5.80
• Adversarial Feeds Steer LLM Agent Decisions Against Their Defaults (2026)
  Rana Muhammad Usman
  5.76
• Constraint Tax in Open-Weight LLMs: An Empirical Study of Tool Calling Suppression Under Structured Output Constraints (2026)
  Fangzheng Li et al.
  5.76
• ProofAgent Harness: Open Infrastructure for Adversarial Evaluation of AI Agents (2026)
  Fouad Bousetouane
  5.71
• LACUNA: Safe Agents as Recursive Program Holes (2026)
  Yaoyu Zhao et al.
  5.71
• Delegated Authorization for Agents Constrained to Semantic Task-to-Scope Matching (2025)
  Majed El Helou et al.
  5.57
• A Synergistic Multi-Agent Framework for Camouflage Attack on Large Language Models (2026)
  Xueyi Li et al.
  5.47
• EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety Guards (2026)
  Siddhant Panpatil et al.
  5.44
• The Illusion of High Utility in Safety Alignment of Text-to-Image Diffusion Models (2026)
  Adeel Yousaf et al.
  5.44
• Robust for the Wrong Reasons: The Representational Geometry of LLM Robustness to Science Skepticism (2026)
  Minjong Cheon
  5.44