![A collage graphic displays various research projects from the 2020 year](https://webarchive.library.unt.edu/web/20201218191745im_/https://www.microsoft.com/en-us/research/uploads/prod/2020/12/1400x788_EOY_blog_stills_no_logo-2-1066x600.jpg)
Microsoft Research
Featured
![A collage graphic displays various research projects from the 2020 year](https://webarchive.library.unt.edu/web/20201218191745im_/https://www.microsoft.com/en-us/research/uploads/prod/2020/12/1400x788_EOY_blog_stills_no_logo-2-1066x600.jpg)
![diagram, schematic](https://webarchive.library.unt.edu/web/20201218191745im_/https://www.microsoft.com/en-us/research/uploads/prod/2020/12/1400x788_RNN_Pool_Still_nologo-5fd3cb7785f56-655x368.jpg)
‘Seeing’ on tiny battery-powered microcontrollers with RNNPool
Learn more about the NeurIPS paper![](https://webarchive.library.unt.edu/web/20201218191745im_/https://www.microsoft.com/en-us/research/uploads/prod/2020/12/1400x788_mpnet_no_logo_still-655x368.jpg)
MPNet combines strengths of masked and permuted language modeling for language understanding
Learn more about the NeurIPS paper![Graphical representation of how our adversarial training approach works for solving the instrumental variable problem. The problem is viewed as a zero-sum game between a learner and an adversary. The learner attempts to find models that satisfy all moment constraints and the adversary flags violating moment constraints. Then the learner tries to correct the model to also satisfy the flagged constraint. A good model is learned when the adversary cannot find large violations](https://webarchive.library.unt.edu/web/20201218191745im_/https://www.microsoft.com/en-us/research/uploads/prod/2020/12/1400x788_Minimax_still_no_logo-1-655x368.jpg)
Adversarial machine learning and instrumental variables for flexible causal modeling
Learn more about the NeurIPS paper![](https://webarchive.library.unt.edu/web/20201218191745im_/https://www.microsoft.com/en-us/research/uploads/prod/2020/11/1400x788_Innereye_still_no_logo-1-655x368.jpg)