Impact Analytics
Measuring how AI research translates into real-world products. These metrics go beyond citations — they track actual product adoption.
144
Researchers
90
Builders
150
Papers
91
Projects
216
Paper→Product Links
37.3%
Translation Rate
Applied Impact Index Leaderboard
Researchers ranked by builder-declared product usage
Reflects builder-declared usage. Not a measure of research quality.
Max Welling
University of Amsterdam / Microsoft Research
Yoshua Bengio
Mila / Université de Montréal
Ilya Sutskever
Safe Superintelligence Inc.
Kaiming He
MIT / Google DeepMind
Fei-Fei Li
Stanford University
Ian Goodfellow
Google DeepMind
Noam Shazeer
Ashish Vaswani
Essential AI
Diederik P. Kingma
Google DeepMind
Jacob Devlin
Luke Zettlemoyer
University of Washington / Meta AI
Thomas Kipf
Google DeepMind
Alec Radford
OpenAI
Phillip Isola
MIT CSAIL
Ali Farhadi
University of Washington / Allen AI
Aleksander Madry
MIT CSAIL
Jure Leskovec
Stanford University
Yarin Gal
University of Oxford
Yann LeCun
Meta AI / NYU
Jürgen Schmidhuber
KAUST / IDSIA
Tomas Mikolov
Czech Institute of Informatics (CIIRC)
Tom Brown
OpenAI
Noah Smith
University of Washington / Allen AI
Hannaneh Hajishirzi
University of Washington / Allen AI
Richard Sutton
University of Alberta / DeepMind
Kyunghyun Cho
New York University / Genentech
Aaron Courville
Mila / Université de Montréal
Andrew Zisserman
University of Oxford
Trevor Darrell
UC Berkeley
Karen Simonyan
DeepMind
Jason Weston
Meta AI
Liang-Chieh Chen
Google DeepMind
Andrej Karpathy
Independent
Yejin Choi
University of Washington / Allen AI
Silvio Savarese
Salesforce AI Research
Hengshuang Zhao
University of Hong Kong
Jeff Dean
Google DeepMind
Bernhard Scholkopf
Max Planck Institute for Intelligent Systems
Stuart Russell
UC Berkeley
Dario Amodei
Anthropic
Saining Xie
New York University
Dawn Song
UC Berkeley
Song Han
MIT
Anima Anandkumar
Caltech / NVIDIA
Michael Bronstein
University of Oxford
Regina Barzilay
MIT CSAIL
Petar Velickovic
Google DeepMind
Been Kim
Google DeepMind
Jiajun Wu
Stanford University
Demis Hassabis
Google DeepMind
Cordelia Schmid
INRIA / Google Research
Most Adopted Papers
Papers ranked by number of products built from them
Attention Is All You Need
NeurIPS 2017 · 2017 · 6,510 citations
products
Language Models are Few-Shot Learners
NeurIPS 2020 · 2020 · 3,027 citations
products
Learning Transferable Visual Models From Natural Language Supervision
ICML 2021 · 2021 · 5,296 citations
products
Adam: A Method for Stochastic Optimization
ICLR 2015 · 2015 · 130,000 citations
products
GPT-4 Technical Report
arXiv preprint · 2023 · 8,500 citations
products
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
NeurIPS 2022 · 2022 · 153 citations
products
ImageNet: A Large-Scale Hierarchical Image Database
CVPR 2009 · 2009 · 60,625 citations
products
Generative Adversarial Nets
NeurIPS 2014 · 2014 · 21,735 citations
products
Deep Residual Learning for Image Recognition
CVPR 2016 · 2015 · 220,000 citations
products
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
NAACL 2019 · 2019 · 108,124 citations
products
Deep contextualized word representations
NAACL 2018 · 2018 · 15,000 citations
products
Towards Deep Learning Models Resistant to Adversarial Attacks
ICLR 2018 · 2018 · 12,000 citations
products
Image-to-Image Translation with Conditional Adversarial Networks
CVPR 2017 · 2017 · 18,000 citations
products
You Only Look Once: Unified, Real-Time Object Detection
CVPR 2016 · 2016 · 35,000 citations
products
Hierarchical Text-Conditional Image Generation with CLIP Latents
arXiv preprint · 2022 · 5,200 citations
products
Training language models to follow instructions with human feedback
NeurIPS 2022 · 2022 · 4,260 citations
products
Efficient Estimation of Word Representations in Vector Space
ICLR Workshop 2013 · 2013 · 38,000 citations
products
Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
EMNLP 2014 · 2014 · 16,000 citations
products
Language Models are Few-Shot Learners
NeurIPS 2020 · 2020 · 18,500 citations
products
Data-Centric Artificial Intelligence: A Survey
Communications of the ACM · 2023 · 1,200 citations
products
Gradient-based learning applied to document recognition
Proceedings of the IEEE · 1998 · 57,014 citations
products
Llama 2: Open Foundation and Fine-Tuned Chat Models
arXiv preprint · 2023 · 7,200 citations
products
OLMo: Accelerating the Science of Language Models
ACL 2024 · 2024 · 800 citations
products
Reinforcement Learning: An Introduction
MIT Press · 2018 · 65,000 citations
products
Semi-Supervised Classification with Graph Convolutional Networks
ICLR 2017 · 2017 · 21,000 citations
products
Long Short-Term Memory
Neural Computation · 1997 · 95,000 citations
products
Liquid Time-constant Networks
AAAI 2021 · 2021 · 249 citations
products
Causal Inference for Statistics, Social, and Biomedical Sciences
Cambridge University Press · 2015 · 5,400 citations
products
Large-scale Video Classification with Convolutional Neural Networks
CVPR 2014 · 2014 · 5,800 citations
products
Neural Architecture Search with Reinforcement Learning
ICLR 2017 · 2017 · 5,100 citations
products
Action Recognition with Trajectory-Pooled Deep-Convolutional Descriptors
CVPR 2015 · 2015 · 1,800 citations
products
Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics
CVPR 2018 · 2018 · 3,800 citations
products
Towards Interpretable Machine Learning: A Survey on Methods and Metrics
Electronics · 2019 · 2,200 citations
products
PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows
ICCV 2019 · 2019 · 650 citations
products
SoK: Eternal War in Memory
IEEE S&P 2013 · 2013 · 1,500 citations
products
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
NeurIPS 2017 · 2017 · 8,500 citations
products
Memory Networks
ICLR 2015 · 2015 · 4,200 citations
products
GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks
ICML 2018 · 2018 · 1,400 citations
products
Variational Graph Auto-Encoders
NeurIPS Workshop 2016 · 2016 · 3,100 citations
products
A Generative Model for Molecular Distance Geometry
ICML 2020 · 2020 · 450 citations
products
COMET: Commonsense Transformers for Automatic Knowledge Graph Construction
ACL 2019 · 2019 · 1,500 citations
products
Very Deep Convolutional Networks for Large-Scale Image Recognition
ICLR 2015 · 2015 · 92,000 citations
products
Auto-Encoding Variational Bayes
ICLR 2014 · 2014 · 22,000 citations
products
Highly accurate protein structure prediction with AlphaFold
Nature · 2021 · 42,952 citations
products
Graph Attention Networks
ICLR 2018 · 2018 · 14,000 citations
products
Concrete Problems in AI Safety
arXiv preprint · 2016 · 1,800 citations
products
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
IEEE TPAMI · 2017 · 14,000 citations
products
Pyramid Scene Parsing Network
CVPR 2017 · 2017 · 9,500 citations
products
Aggregated Residual Transformations for Deep Neural Networks
CVPR 2017 · 2017 · 10,500 citations
products
Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
ICLR 2016 · 2016 · 7,200 citations
products
Geometric Deep Learning: Going Beyond Euclidean Data
IEEE Signal Processing Magazine · 2017 · 4,500 citations
products
Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
ICML 2016 · 2016 · 8,200 citations
products
Representation Learning on Graphs: Methods and Applications
IEEE Data Engineering Bulletin · 2017 · 3,200 citations
products
Artificial Intelligence: A Modern Approach
Pearson (4th Edition) · 2020 · 72,000 citations
products
High-Resolution Image Synthesis with Latent Diffusion Models
CVPR 2022 · 2022 · 14,000 citations
products
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
ICLR 2019 · 2019 · 6,500 citations
products
Research → Product Domain Flow
How research domains translate into product domains
Domain Breakdown
Research and product activity by domain
| Domain | Researchers | Papers | Projects |
|---|---|---|---|
| Generative AI | 51 | 27 | 77 |
| NLP | 53 | 26 | 58 |
| Computer Vision | 46 | 27 | 27 |
| AI Safety | 23 | 11 | 2 |
| Optimization | 17 | 10 | 4 |
| Robotics | 12 | 7 | 7 |
| Reinforcement Learning | 12 | 6 | 4 |
| Multimodal | 8 | 3 | 5 |
| MLOps | 4 | 3 | 5 |
| Healthcare AI | 8 | 3 | 1 |
| Other | 7 | 4 | 0 |
| Deep Learning | 6 | 7 | 0 |
| Machine Learning | 6 | 3 | 0 |
| Graph ML | 5 | 5 | 0 |
| Machine Learning Theory | 5 | 3 | 0 |
| Computational Linguistics | 3 | 1 | 0 |
| Audio AI | 0 | 0 | 3 |
| Knowledge Graphs | 2 | 0 | 0 |
| AI Fairness | 2 | 1 | 0 |
| Recurrent Networks | 1 | 1 | 0 |
| Explainable AI | 1 | 1 | 0 |
| Object Tracking | 1 | 1 | 0 |
| Scene Understanding | 1 | 1 | 0 |
| Transfer Learning | 1 | 1 | 0 |
| Data Mining | 1 | 1 | 0 |
| Social Networks | 1 | 1 | 0 |
| Information Retrieval | 1 | 1 | 0 |
| Robustness | 1 | 0 | 0 |
| Continual Learning | 1 | 0 | 0 |
| Question Answering | 1 | 1 | 0 |
| Action Recognition | 1 | 0 | 0 |
| Summarization | 1 | 0 | 0 |
| Self-Supervised Learning | 1 | 1 | 0 |
| Graph Neural Networks | 1 | 1 | 0 |
| Argumentation Mining | 1 | 0 | 0 |
| Knowledge Bases | 1 | 1 | 0 |
| Online Learning | 1 | 1 | 0 |
| Conversational AI | 1 | 0 | 0 |
| Data-Centric AI | 1 | 1 | 0 |
| Machine Learning Systems | 1 | 1 | 0 |
| Computational Social Science | 1 | 0 | 0 |
| Parsing | 1 | 0 | 0 |
| Dialogue Systems | 1 | 0 | 0 |
| Fact Verification | 1 | 1 | 0 |
| Low-Resource Languages | 1 | 0 | 0 |
| Human-Robot Interaction | 1 | 0 | 0 |
| Reasoning | 1 | 0 | 0 |
| Speech & Audio | 0 | 0 | 1 |
Domain Leaderboards
Top 3 researchers by Applied Impact Index per domain
Paper→product link data is CC0 licensed. Free to use, share, and build on.
About These Metrics — Applied Impact Index v1.0.0 — measures real-world product adoption of research papers.
Version: 1.0.0
The Applied Impact Index (AII) measures real-world product adoption of academic research based on builder-declared usage. Unlike citation-based metrics (h-index, impact factor) or attention-based metrics (Altmetric score), AII tracks how many products and projects were actually built using a researcher's work. The score combines four components: Product Adoption Count (40%), Domain Breadth (30%), Foundation Index (20%), and Translation Rate (10%). The Translation Rate measures what percentage of papers have at least one real-world product — a metric no existing platform (Altmetric, PlumX, Overton, Dimensions, Lens.org, or Google Scholar) currently provides.
Disclaimer: Scores reflect builder-declared project usage. They are not a measure of research quality. Scoring methodology is open source and reproducible. Data is licensed CC0.