Scientific Paper · 2019-06-05
Energy and Policy Considerations for Deep Learning in NLP
Training large neural networks improves NLP accuracy but requires significant computational resources, leading to high financial and environmental costs.
ENVIRONMENTAL TOPIC
Energy and electricity
Electricity supply, demand, grids, fuels, and energy efficiency.
Browse all 836 resources →Counts reflect provisional metadata classification, not measured environmental impact. A record may belong to multiple topics.
View this topic through four lenses
Share of the 836 records in this view. Pathways can overlap; bars do not add to 100%.
Explore the connected collections
Research papers and scholarly literature on impacts, applications, and methods.
Scientific Paper · 2024-10-15
Power Hungry Processing: Watts Driving the Cost of AI Deployment?
Generative AI models consume significantly more energy and emit more carbon than task-specific models, raising concerns about their environmental impact.
Scientific Paper · 2024-03-27
The Climate and Sustainability Implications of Generative AI
The unchecked growth of generative AI increases energy demand and environmental impact, requiring a sustainability-focused approach beyond efficiency improvements.
Scientific Paper · 2024-12-04
A Deep Learning Based Framework to Identify Undocumented Orphaned Oil and Gas Wells from Historical Maps: A Case Study for California and Oklahoma
A neural network model identifies undocumented orphaned wells, helping locate methane-leaking sites and aiding environmental monitoring efforts.
What is missing from the metadata?
Missing metadata is a curation task, not evidence of absent research or activity. The Resource Hub does not currently contain a validated incident dataset, expert survey, or measurements of net environmental benefit.