Date: 10 May 2023
The transformer architecture has remarkably improved the accuracy of sequence modeling tasks (eg. language translation) but could they be used to model chemical reactions? Do chemical reactions have some similarity with natural language that allows for the use of powerful transformer-inspired architectures? How would the molecules be represented in such...
Date: 20 April 2023
Imagine we create a huge, single network of organic chemistry reactions where each unique molecule is represented as a node and molecules that take part in a reaction are connected with directed edges. Such a network would be representative of the current reaction chemistry knowledge. Obviously, some nodes would be...
Date: 11 April 2023
Ever wondered how to inject chemistry knowledge into machine learning models to accurately predict properties of molecules? How would a molecule be represented before it is input to an ML model? How could the representation be made more chemistry-aware and ensure it is complete and accurate? Do natural language processing...
Date: 12 March 2021
We have all been hearing the following auguries quite frequently in recent years -- "Machines are going to replace humans in all the tasks that you could imagine?", "Artificial Intelligence is going to solve all the problems (but invent several others) that the human civilization currently suffers from!" and my...
Date: 9 July 2019
As the academic year comes to a close, a whole new generation of workforce is getting ready to enter the hyper-competitive job market. The constituents of this workforce are largely people who have worked really hard over the last few years and have been successful in securing a job right...