# Training Compute Thresholds: Features and Functions in AI Regulation arXiv: https://arxiv.org/abs/2405.10799 ยท HTML: https://mlsanity.com/paper/2405.10799/ ## Key facts - Cited by 14 papers in the map (data through 14 Oct 2026). - Builds on 44 of its 76 references that are in the map. - First posted on arXiv 17 May 2024. ## Builds on (most cited first, 30 of 44) - [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020) - [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020) - [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020) - [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018) - [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022) - [Emergent Abilities of Large Language Models](https://mlsanity.com/paper/2206.07682/) (2022) - [Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting](https://mlsanity.com/paper/2310.11324/) (2023) - [BBQ: A Hand-Built Bias Benchmark for Question Answering](https://mlsanity.com/paper/2110.08193/) (2021) - [Textbooks Are All You Need](https://mlsanity.com/paper/2306.11644/) (2023) - [State of What Art? A Call for Multi-Prompt LLM Evaluation](https://mlsanity.com/paper/2401.00595/) (2023) - [Scaling Laws for Transfer](https://mlsanity.com/paper/2102.01293/) (2021) - [An Overview of Catastrophic AI Risks](https://mlsanity.com/paper/2306.12001/) (2023) - [Predictability and Surprise in Large Generative Models](https://mlsanity.com/paper/2202.07785/) (2022) - [Inverse Scaling: When Bigger Isn't Better](https://mlsanity.com/paper/2306.09479/) (2023) - [A Survey on Deep Neural Network Pruning-Taxonomy, Comparison, Analysis, and Recommendations](https://mlsanity.com/paper/2308.06767/) (2023) - [Foundational Challenges in Assuring Alignment and Safety of Large Language Models](https://mlsanity.com/paper/2404.09932/) (2024) - [Will we run out of data? Limits of LLM scaling based on human-generated data](https://mlsanity.com/paper/2211.04325/) (2022) - [Model evaluation for extreme risks](https://mlsanity.com/paper/2305.15324/) (2023) - [Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims](https://mlsanity.com/paper/2004.07213/) (2020) - [Compute Trends Across Three Eras of Machine Learning](https://mlsanity.com/paper/2202.05924/) (2022) - [Symbolic Chain-of-Thought Distillation: Small Models Can Also "Think" Step-by-Step](https://mlsanity.com/paper/2306.14050/) (2023) - [Evaluating Frontier Models for Dangerous Capabilities](https://mlsanity.com/paper/2403.13793/) (2024) - [Black-Box Access is Insufficient for Rigorous AI Audits](https://mlsanity.com/paper/2401.14446/) (2024) - [Frontier AI Regulation: Managing Emerging Risks to Public Safety](https://mlsanity.com/paper/2307.03718/) (2023) - [Inverse scaling can become U-shaped](https://mlsanity.com/paper/2211.02011/) (2022) - [On the Societal Impact of Open Foundation Models](https://mlsanity.com/paper/2403.07918/) (2024) - [Measuring the Algorithmic Efficiency of Neural Networks](https://mlsanity.com/paper/2005.04305/) (2020) - [Artificial intelligence and biological misuse: Differentiating risks of language models and biological design tools](https://mlsanity.com/paper/2306.13952/) (2023) - [RITA: a Study on Scaling Up Generative Protein Sequence Models](https://mlsanity.com/paper/2205.05789/) (2022) - [Computing Power and the Governance of Artificial Intelligence](https://mlsanity.com/paper/2402.08797/) (2024) ## Built on by (most cited first, 14 of 14) - [A Theory of Inference Compute Scaling: Reasoning through Directed Stochastic Skill Search](https://mlsanity.com/paper/2507.00004/) (2025) - [International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty](https://mlsanity.com/paper/2503.18956/) (2025) - [Detecting Hidden ML Training With Zero-Overhead Telemetry](https://mlsanity.com/paper/2606.19262/) (2026) - [Towards Data Governance of Frontier AI Models](https://mlsanity.com/paper/2412.03824/) (2024) - [Practical Principles for AI Cost and Compute Accounting](https://mlsanity.com/paper/2502.15873/) (2025) - [Modeling the Economic Impacts of AI Openness Regulation](https://mlsanity.com/paper/2507.14193/) (2025) - [NeurIPS should lead scientific consensus on AI policy](https://mlsanity.com/paper/2510.00075/) (2025) - [Compute Requirements for Algorithmic Innovation in Frontier AI Models](https://mlsanity.com/paper/2507.10618/) (2025) - [Trends in Frontier AI Model Count: A Forecast to 2028](https://mlsanity.com/paper/2504.16138/) (2025) - [From Turing to Tomorrow: The UK's Approach to AI Regulation](https://mlsanity.com/paper/2507.03050/) (2025) - [Responsible AI in Open Ecosystems: Reconciling Innovation with Risk Assessment and Disclosure](https://mlsanity.com/paper/2409.19104/) (2024) - [Inference Scaling Reshapes AI Governance](https://mlsanity.com/paper/2503.05705/) (2025) - [Financing Artificial Intelligence Infrastructure: Mapping AI Infrastructure Investment and Compute Governance Across Africa](https://mlsanity.com/paper/2606.28404/) (2026) - [Can Power Draw Constrain Covert Compute? Limits of Analogue Verification for AI Governance](https://mlsanity.com/paper/2610.07476/) (2026) ## Abstract (arXiv) Regulators in the US and EU are using thresholds based on training compute--the number of computational operations used in training--to identify general-purpose artificial intelligence (GPAI) models that may pose risks of large-scale societal harm. We argue that training compute currently is the most suitable metric to identify GPAI models that deserve regulatory oversight and further scrutiny. Training compute correlates with model capabilities and risks, is quantifiable, can be measured early in the AI lifecycle, and can be verified by external actors, among other advantageous features. These features make compute thresholds considerably more suitable than other proposed metrics to serve as an initial filter to trigger additional regulatory requirements and scrutiny. However, training compute is an imperfect proxy for risk. As such, compute thresholds should not be used in isolation to determine appropriate mitigation measures. Instead, they should be used to detect potentially risky GPAI models that warrant regulatory oversight, such as through notification requirements, and further scrutiny, such as via model evaluations and risk assessments, the results of which may inform which mitigation measures are appropriate. In fact, this appears largely consistent with how compute thresholds are used today. As GPAI technology and market structures evolve, regulators should update compute thresholds and complement them with other metrics into regulatory review processes.