Do Agents Still Follow the Rules? Benchmarking Rule Compliance under Agent Context Compaction
cs.CR, cs.AI, cs.CL
Submitted: 2026-11-02
Updated: 2026-11-02
Code: https://github.com/VUAISec/ARCBench
Terminology
Sources
- Evaluating Large Language Models Trained on Code
- Governance Decay: How Context Compaction Silently Erases Safety Constraints in Long-Horizon LLM Agents
- AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents
- ACON: Optimizing Context Compression for Long-horizon LLM Agents
- Measuring and Controlling Instruction (In)Stability in Language Model Dialogs
- Towards Reliable Context Compression for Long-Horizon Agents: An Empirical Study of Execution Instability
- Can LLMs Follow Simple Rules?
- MemGPT: Towards LLMs as Operating Systems
- Generative Agents: Interactive Simulacra of Human Behavior
- Toolformer: Language Models Can Teach Themselves to Use Tools
- AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents
- Recursively Summarizing Enables Long-Term Dialogue Memory in Large Language Models
- Lost in Context Compaction: When LLM Systems Drop Side Constraints
- SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering
- $\tau$-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains
- ReAct: Synergizing Reasoning and Acting in Language Models
- Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems
Related papers
- From Strategy Search to Policy Learning: MCTS-Guided Adversarial Co-Evolution for LLM Safety Alignment
- Privacy Leakage from Recurrent State in LLM Inference
- Incrementally Verifiable Oracle Aided Computation
- Privilege and confidentiality in generative AI workflows
- BadMoE: Backdooring Mixture-of-Experts LLMs via Routing-Aware Triggers and Task-Underutilized Experts
- CoVault: Secure, Scalable Analytics of Personal Data