Generative modeling methods
A research field
Generative modeling methods is losing ground in AI research: 2.3% to 2.0% of new AI papers, -0.3 points. It ranks #39 of 44 by gain in share. As of Oct 10, 2026.
- Change in share
- -0.3 pts
- Share, last 3 months
- 2.0%
- Papers, last 3 months
- 661
- Ready for products
- 2.4 / 5
Ideas in this field
Flow matching0.2% → 0.3%+0.0 ptsGenerative adversarial network0.0% → 0.0%0.0 ptsDiffusion guidance0.1% → 0.1%0.0 ptsConditional variational autoencoder0.0% → 0.0%0.0 ptsDiffusion sampling0.2% → 0.1%0.0 ptsNormalizing flow0.0% → 0.0%0.0 ptsEnergy based model0.0% → 0.0%0.0 ptsConditional flow matching0.0% → 0.0%0.0 ptsControllable text generation0.0% → 0.0%0.0 ptsRectified flow0.0% → 0.0%0.0 pts
Companies working on it
Generative modeling methods: quick answers
No. Its share of new AI papers went from 2.3% to 2.0%, with 661 papers in the last 3 months.