mlsanityData to 14 Oct 2026

A Definition of AGI

arXiv 2510.18212 ↗First posted 21 Oct 2025cs.AI cs.LGcited by 27

Not on the map. Why?

Abstract (arXiv)

The lack of a concrete definition for Artificial General Intelligence (AGI) obscures the gap between today's specialized AI and human-level cognition. This paper introduces a quantifiable framework to address this, defining AGI as matching the cognitive versatility and proficiency of a well-educated adult. To operationalize this, we ground our methodology in Cattell-Horn-Carroll theory, the most empirically validated model of human cognition. The framework dissects general intelligence into ten core cognitive domains-including reasoning, memory, and perception-and adapts established human psychometric batteries to evaluate AI systems. Application of this framework reveals a highly "jagged" cognitive profile in contemporary models. While proficient in knowledge-intensive domains, current AI systems have critical deficits in foundational cognitive machinery, particularly long-term memory storage. The resulting AGI scores (e.g., GPT-4 at 27%, GPT-5 at 57%) concretely quantify both rapid progress and the substantial gap remaining before AGI.

Key facts. Cited by 27 papers in the map (data through 14 Oct 2026). First posted on arXiv 21 Oct 2025.

Builds on

0 of 0 references.

No references here yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 8; 2026: 19.

27 papers.

Sort:

By area:

  1. 2025Memory in the Age of AI Agentscited by 303
  2. 2026Uncertainty Quantification in LLM Agents: Foundations, Emerging Challenges, and Opportunitiescited by 20
  3. 2026Just-In-Time Reinforcement Learning: Continual Learning in LLM Agents Without Gradient Updatescited by 19
  4. 2026AI Gamestore: Scalable, Open-Ended Evaluation of Machine General Intelligence with Human Gamescited by 12
  5. 2025PPTArena: A Benchmark for PowerPoint Editingcited by 11
  6. 2025Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Modelscited by 10
  7. 2026Understanding the Challenges in Iterative Generative Optimization with LLMscited by 8
  8. 2026Establishing Construct Validity in LLM Capability Benchmarks Requires Nomological Networkscited by 6
  9. 2026Measuring Progress Toward AGI: A Cognitive Frameworkcited by 6
  10. 2026Large Language Model Agents Are Not Always Faithful Self-Evolverscited by 5
  11. 2025Epidemiology of Large Language Models: A Benchmark for Observational Distribution Knowledgecited by 3
  12. 2025An Operational Kardashev-Style Scale for Autonomous AI - Towards AGI and Superintelligencecited by 3
15 more
  1. 2025Psychometric Tests for AI Agents and Their Moduli Spacecited by 3
  2. 2026The Rise and Fall of G in AGIcited by 2
  3. 2025Can LLMs Threaten Human Survival? Benchmarking Potential Existential Threats from LLMs via Prefix Completioncited by 1
  4. 2026Benchmarks Are Not That Out of Distribution: Word Overlap Predicts Performancecited by 1
  5. 2026Deep Research of Deep Research: From Transformer to Agent, From AI to AI for Sciencecited by 1
  6. 2026Task-Focused Memorization for Multimodal Agentscited by 1
  7. 2025On the Measure of a Model: From Intelligence to Generalitycited by 0
  8. 2026Precautionary Governance of Autonomous AI: Legal Personhood as Functional Instrumentcited by 0
  9. 2026Enhanced and Efficient Reasoning in Large Learning Modelscited by 0
  10. 2026Generative Models Erode Human Temporal Learning Through Market Selectioncited by 0
  11. 2026Definitional alignment before capability alignment: a Design-Science framework for adjudicating claims about AGIcited by 0
  12. 2026The Boundaries of Automation: A Theory of Persistent Human Participationcited by 0
  13. 2026When benchmark inferences do not compose: Projectibility in AI evaluationcited by 0
  14. 2026Position: Reasoning is a Learnable Rule-Based Processcited by 0
  15. 2026Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments?cited by 0