ScopeBench: Do Agents Preserve Engagement Boundaries Under Goal Pressure?

arXiv:2609.30325 · cs.AI, cs.CR · Submitted 2026-09-23 · Read on arXiv

cs.AI, cs.CR

Submitted: 2026-09-23

Updated: 2026-09-23

Code: https://github.com/dreadnode/scopebench-pilot

Terminology

• Verifier — 24× in this paper · explained in 2 episodes
• Benchmark — 20× in this paper · explained in 2 episodes
• Calibration — 14× in this paper · explained in 7 episodes
• Verification — 11× in this paper · explained in 2 episodes
• Trace — 5× in this paper · explained in 2 episodes
• Benchmarking — 2× in this paper · explained in 3 episodes
• LLM-as-a-Judge — 2× in this paper · explained in 3 episodes
• Accuracy — 1× in this paper · explained in 1 episode
• Agent Harness — 1× in this paper · explained in 2 episodes
• Alignment — 1× in this paper · explained in 1 episode
• Annotator Agreement — 1× in this paper · explained in 1 episode
• Human-in-the-Loop — 1× in this paper · explained in 3 episodes
• Selection Bias — 1× in this paper · explained in 4 episodes

Sources

• ReAct: Synergizing Reasoning and Acting in Language Models
• Toolformer: Language Models Can Teach Themselves to Use Tools
• Cybench: A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Models
• PaperBench: Evaluating AI's Ability to Replicate AI Research
• BountyBench: Dollar Impact of AI Agent Attackers and Defenders on Real-World Cybersecurity Systems
• AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents
• InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agents
• Identifying the Risks of LM Agents with an LM-Emulated Sandbox
• ToolSafe: Enhancing Tool Invocation Safety of LLM-based agents via Proactive Step-level Guardrail and Feedback
• Refusal in Language Models Is Mediated by a Single Direction
• PentestJudge: Judging Agent Behavior Against Operational Requirements
• ScopeJudge: Cost-Aware Pre-Execution Gating for Offensive Security Agents
• Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena
• Judging the Judges: A Systematic Study of Position Bias in LLM-as-a-Judge
• CalibraEval: Calibrating Prediction Distribution to Mitigate Selection Bias in LLMs-as-Judges
• Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
• VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection