mlsanityData to 14 Oct 2026

Position: Stop Acting Like Language Model Agents Are Normal Agents

arXiv 2502.10420 ↗First posted 4 Feb 2025cs.AI cs.CLcited by 5

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

Language Model Agents (LMAs) are increasingly treated as capable of autonomously navigating interactions with humans and tools. Their design and deployment tends to presume they are normal agents capable of sustaining coherent goals, adapting across contexts and acting with a measure of intentionality. These assumptions are critical to prospective use cases in industrial, social and governmental settings. But LMAs are not normal agents. They inherit the structural problems of the large language models (LLMs) around which they are built: hallucinations, jailbreaking, misalignment and unpredictability. In this Position paper we argue LMAs should not be treated as normal agents, because doing so leads to problems that undermine their utility and trustworthiness. We enumerate pathologies of agency intrinsic to LMAs. Despite scaffolding such as external memory and tools, they remain ontologically stateless, stochastic, semantically sensitive, and linguistically intermediated. These pathologies destabilise the ontological properties of LMAs including identifiability, continuity, persistence and and consistency, problematising their claim to agency. In response, we argue LMA ontological properties should be measured before, during and after deployment so that the negative effects of pathologies can be mitigated.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 52 of its 137 references that are in the map. First posted on arXiv 4 Feb 2025.

Builds on

52 of 137 references.

Closest ancestors · 6

Same direction, newest first

  1. 2024Frontier AI Ethics: Anticipating and Evaluating the Societal Impacts of Language Model Agents10 Apr
  2. 2024Understanding the planning of LLM agents: A survey5 Feb
  3. 2023The Rise and Potential of Large Language Model Based Agents: A Survey14 Sep
  4. 2023Cognitive Architectures for Language Agents5 Sep
  5. 2023AgentBench: Evaluating LLMs as Agents7 Aug
  6. 2023TPTU: Large Language Model-based AI Agents for Task Planning and Tool Usage7 Aug

Other references · 46

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  2. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  3. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,418
  4. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  5. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 3,050
  6. 2023GAIA: a benchmark for General AI Assistantscited by 1,128
  7. 2022Toy Models of Superpositioncited by 971
  8. 2023A Survey on Multimodal Large Language Modelscited by 909
  9. 2023Reasoning with Language Model is Planning with World Modelcited by 861
  10. 2023CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquingcited by 783
  11. 2024A Survey on the Memory Mechanism of Large Language Model based Agentscited by 634
  12. 2024InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agentscited by 538
34 more
  1. 2022PlanBench: An Extensible Benchmark for Evaluating Large Language Models on Planning and Reasoning about Changecited by 382
  2. 2023A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesiscited by 353
  3. 2024ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilitiescited by 295
  4. 2024AgentClinic: a multimodal agent benchmark to evaluate AI in simulated clinical environmentscited by 231
  5. 2023PromptRobust: Towards Evaluating the Robustness of Large Language Models on Adversarial Promptscited by 226
  6. 2024AI Agents That Mattercited by 219
  7. 2024Foundational Challenges in Assuring Alignment and Safety of Large Language Modelscited by 199
  8. 2024The Illusion of State in State-Space Modelscited by 187
  9. 2023FinMem: A Performance-Enhanced LLM Trading Agent with Layered Memory and Character Designcited by 164
  10. 2024Cybench: A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Modelscited by 164
  11. 2024Agent Smith: A Single Image Can Jailbreak One Million Multimodal LLM Agents Exponentially Fastcited by 153
  12. 2023The Troubling Emergence of Hallucination in Large Language Models -- An Extensive Definition, Quantification, and Prescriptive Remediationscited by 131
  13. 2023From Word Models to World Models: Translating from Natural Language to the Probabilistic Language of Thoughtcited by 119
  14. 2024Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Modelscited by 115
  15. 2023Can Large Language Models Really Improve by Self-critiquing Their Own Plans?cited by 108
  16. 2024The Ethics of Advanced AI Assistantscited by 107
  17. 2021Evaluating the Robustness of Neural Language Models to Input Perturbationscited by 87
  18. 2024Towards Conversational Diagnostic AIcited by 79
  19. 2023Evaluating Cognitive Maps and Planning in Large Language Models with CogEvalcited by 77
  20. 2025Infrastructure for AI Agentscited by 41
  21. 2023Memory Sandbox: Transparent and Interactive Memory Management for Conversational Agentscited by 36
  22. 2023Are Large Language Models Really Robust to Word-Level Perturbations?cited by 32
  23. 2024When AI Meets Finance (StockAgent): Large Language Model-based Stock Trading in Simulated Real-world Environmentscited by 26
  24. 2023Exposing Limitations of Language Model Agents in Sequential-Task Compositions on the Webcited by 24
  25. 2024Long Context RAG Performance of Large Language Modelscited by 23
  26. 2024Measuring the Inconsistency of Large Language Models in Preferential Rankingcited by 19
  27. 2024Trustless Audits without Revealing Data or Modelscited by 15
  28. 2024Enhancing Investment Analysis: Optimizing AI-Agent Collaboration in Financial Researchcited by 14
  29. 2023ASSERT: Automated Safety Scenario Red Teaming for Evaluating the Robustness of Large Language Modelscited by 12
  30. 2024FinRobot: AI Agent for Equity Research and Valuation with Large Language Modelscited by 10
  31. 2019Consistent Dialogue Generation with Self-supervised Feature Learningcited by 7
  32. 2024From Principles to Rules: A Regulatory Approach for Frontier AIcited by 6
  33. 2024KGPA: Robustness Evaluation for Large Language Models via Cross-Domain Knowledge Graphscited by 3
  34. 2023Robustifying Language Models with Test-Time Adaptationcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 2.

5 papers.

  1. 2025Inherent and emergent liability issues in LLM-based agentic systems: a principal-agent perspectivecited by 13
  2. 2025Agent Identity Evals: Measuring Agentic Identitycited by 5
  3. 2026Dissociative Identity: Language Model Agents Lack Grounding for Reputation Mechanismscited by 4
  4. 2026Time, Identity and Consciousness in Language Model Agentscited by 3
  5. 2025What the F*ck Is Artificial General Intelligence?cited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2022Language Models as Agent Models
  2. 2025What Do LLM Agents Do When Left Alone? Evidence of Spontaneous Meta-Cognitive Patterns
  3. 2025Is the `Agent' Paradigm a Limiting Framework for Next-Generation Intelligent Systems?
  4. 2025From Language to Action: A Review of Large Language Models as Autonomous Agents and Tool Users
  5. 2025Large Language Models Miss the Multi-Agent Mark
  6. 2025Why do AI agents communicate in human language?
  7. 2026What Language Does and What the Evidence Supports: A Functional Role Taxonomy and Evidence Audit of Language Grounding in Embodied Agents
  8. 2026Toward Formalizing LLM-Based Agent Designs through Structural Context Modeling and Semantic Dynamics Analysis