mlsanityData to 7 Oct 2026

Counting Carbon: A Survey of Factors Influencing the Emissions of Machine Learning

arXiv 2302.08476 ↗First posted 16 Feb 2023cs.LG cs.CYcited by 42

Not on the map. Why?

Abstract (arXiv)

Machine learning (ML) requires using energy to carry out computations during the model training process. The generation of this energy comes with an environmental cost in terms of greenhouse gas emissions, depending on quantity used and the energy source. Existing research on the environmental impacts of ML has been limited to analyses covering a small number of models and does not adequately represent the diversity of ML models and tasks. In the current study, we present a survey of the carbon emissions of 95 ML models across time and different tasks in natural language processing and computer vision. We analyze them in terms of the energy sources used, the amount of CO2 emissions produced, how these emissions evolve across time and how they relate to model performance. We conclude with a discussion regarding the carbon footprint of our field and propose the creation of a centralized repository for reporting and tracking these emissions.

Key facts. Cited by 42 papers in the map (data through 7 Oct 2026). Builds on 34 of its 62 references that are in the map. First posted on arXiv 16 Feb 2023.

Builds on

34 of 62 references.

References · 34

  1. 2017Attention Is All You Needcited by 59,932
  2. 2020Language Models are Few-Shot Learnerscited by 32,199
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  4. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,243
  5. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  6. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
  7. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,464
  8. 2022Training Compute-Optimal Large Language Modelscited by 2,927
  9. 2016Neural Architecture Search with Reinforcement Learningcited by 1,489
  10. 2021Multitask Prompted Training Enables Zero-Shot Task Generalizationcited by 1,462
  11. 2019Energy and Policy Considerations for Deep Learning in NLPcited by 1,242
  12. 2019Green AIcited by 620
22 more
  1. 2021Carbon Emissions and Large Neural Network Trainingcited by 561
  2. 2019Quantifying the Carbon Emissions of Machine Learningcited by 516
  3. 2022Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Modelcited by 264
  4. 2020Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learningcited by 246
  5. 2021Sustainable AI: Environmental Implications, Challenges and Opportunitiescited by 242
  6. 2020The Computational Limits of Deep Learningcited by 188
  7. 2022The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrinkcited by 183
  8. 2021The Values Encoded in Machine Learning Researchcited by 172
  9. 2020Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Modelscited by 168
  10. 2020Characterising Bias in Compressed Modelscited by 162
  11. 2022Measuring the Carbon Intensity of AI in Cloud Instancescited by 110
  12. 2020Chasing Carbon: The Elusive Environmental Footprint of Computingcited by 71
  13. 2020The De-democratization of AI: Deep Learning and the Compute Divide in Artificial Intelligence Researchcited by 67
  14. 2021The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translationcited by 46
  15. 2019Energy Usage Reports: Environmental awareness as part of algorithmic accountabilitycited by 43
  16. 2020The Grey Hoodie Project: Big Tobacco, Big Tech, and the threat on academic integritycited by 38
  17. 2021Unraveling the Hidden Environmental Impacts of AI Solutions for Environmentcited by 32
  18. 2022Can Model Compression Improve NLP Fairnesscited by 18
  19. 2020HULK: An Energy Efficiency Benchmark Platform for Responsible Natural Language Processingcited by 16
  20. 2021Towards Quantifying the Carbon Emissions of Differentially Private Machine Learningcited by 8
  21. 2022Bridging Fairness and Environmental Sustainability in Natural Language Processingcited by 5
  22. 2021Curb Your Carbon Emissions: Benchmarking Carbon Emissions in Machine Translationcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 5; 2024: 17; 2025: 13; 2026: 7.

42 papers.

Sort:

By area:

  1. 2023Power Hungry Processing: Watts Driving the Cost of AI Deployment?cited by 186
  2. 2024ALLaM: Large Language Models for Arabic and Englishcited by 55
  3. 2023Ecosystem Graphs: The Social Footprint of Foundation Modelscited by 31
  4. 2025Measuring the environmental impact of delivering AI at Google Scalecited by 31
  5. 2024Towards Trustworthy AI: A Review of Ethical and Robust Large Language Modelscited by 21
  6. 2024The 2024 Foundation Model Transparency Indexcited by 18
  7. 2024Can I trust my fake data -- A comprehensive quality assessment framework for synthetic tabular data in healthcarecited by 17
  8. 2025Virtual Agent Economiescited by 17
  9. 2024Green AI: Exploring Carbon Footprints, Mitigation Strategies, and Trade Offs in Large Language Model Trainingcited by 15
  10. 2023Efficiency is Not Enough: A Critical Perspective of Environmentally Sustainable AIcited by 12
  11. 2024Multimodal Learned Sparse Retrieval with Probabilistic Expansion Controlcited by 12
  12. 2024Towards Sustainable Large Language Model Servingcited by 12
30 more
  1. 2024Position: An Inner Interpretability Framework for AI Inspired by Lessons from Cognitive Neurosciencecited by 11
  2. 2025Towards Sustainable NLP: Insights from Benchmarking Inference Energy in Large Language Modelscited by 11
  3. 2024Finding Lottery Tickets in Vision Models via Data-driven Spectral Foresight Pruningcited by 10
  4. 2025Life-Cycle Emissions of AI Hardware: A Cradle-To-Grave Approach and Generational Trendscited by 10
  5. 2025The 2025 Foundation Model Transparency Indexcited by 9
  6. 2024Hardware Scaling Trends and Diminishing Returns in Large-Scale Distributed Trainingcited by 7
  7. 2024Green My LLM: Studying the key factors affecting the energy consumption of code assistantscited by 6
  8. 2026Beyond Test-Time Compute Strategies: Advocating Energy-per-Token in LLM Inferencecited by 5
  9. 2023Estimating Deep Learning energy consumption based on model architecture and training environmentcited by 4
  10. 2024Training Green AI Models Using Elite Samplescited by 4
  11. 2024AI, Climate, and Regulation: From Data Centers to the AI Actcited by 4
  12. 2023The Ethics of AI Value Chainscited by 3
  13. 2025Climate Implications of Diffusion-based Generative Visual AI Systems and their Mass Adoptioncited by 3
  14. 2025Limits to AI Growth: The Ecological and Social Consequences of Scalingcited by 2
  15. 2025Position: Neglecting the Sustainability of AI is Fuelling a Global AI Arms Racecited by 2
  16. 2024Towards Equitable Agile Research and Development of AI and Roboticscited by 1
  17. 2024Towards Sustainable SecureML: Quantifying Carbon Footprint of Adversarial Machine Learningcited by 1
  18. 2024SeagrassFinder: Deep Learning for Eelgrass Detection and Coverage Estimation in the Wildcited by 1
  19. 2024Social Science Is Necessary for Operationalizing Socially Responsible Foundation Modelscited by 1
  20. 2025Comparing energy consumption and accuracy in text classification inferencecited by 1
  21. 2025Irresponsible AI: big tech's influence on AI research and associated impactscited by 1
  22. 2026CarbonEdge: Carbon-Aware Deep Learning Inference Framework for Sustainable Edge Computingcited by 1
  23. 2026Representational Harms in LLM-Generated Narratives Against Global Majority Nationalitiescited by 1
  24. 2025Local Differences, Global Lessons: Insights from Organisation Policies for International Legislationcited by 0
  25. 2025Digital Overconsumption and Waste: A Closer Look at the Impacts of Generative AIcited by 0
  26. 2025Responsible Data Stewardship: Generative AI and the Digital Waste Problemcited by 0
  27. 2026Responsible AI for General-Purpose Systems: Overview, Challenges, and A Path Forwardcited by 0
  28. 2026Green Physics-Informed Machine Learning Models For Structural Health Monitoringcited by 0
  29. 2026MedicalRec: Medical recommender system for image classification without retrainingcited by 0
  30. 2026Environmental Slow AI: Design Principles for Generative Systemscited by 0