I've been working with flow matching models for video generation for a while, and recently went back to my old notes from when I was first learning about them. I cleaned them up and turned them into this blog post.
Hopefully it’s useful for anyone exploring flow matching for generative modeling. Writing it certainly helped solidify my own understanding.
I build a large-language model based "agent" that can can execute small tasks by using a Python REPL and pre-implemented methods. The "agent loop" is based on ReAct.
The agent can look up methods using a `method_search()` method that uses semantic search on the indexed Python methods.
Yes, and other people I know don't have the same problem. Making me think that everyone who has this issue is part of some A/B testing [1] bucket for an internal Gmail spam filtering experiment.
These A/B tests are quite common for products like these to test new features, such as a new spam filtering algorithm. Or even test the efficiency of an existing spam filtering algorithm by degrading the experience of a subset of users.
Blue Prism | Research Scientists & Engineers | London, UK | Full-time | Onsite
At Blue Prism we developed Robotic Process Automation software to provide businesses and organisations with a more agile virtual workforce. Our software platform enables business operations to be agile and cost effective through rapid automation of manual, rules based, back office administrative processes, reducing cost and improving accuracy by creating a “digital workforce”.
To invest in the future of automation we are building up a machine learning focussed research team in London hiring for researchers and engineers:
Bain & Company | Machine Learning Engineer | London, UK / San Francisco, US / Boston, US | onsite + Travel | Full Time
As a top management consulting firm, Bain & Co helps the world’s top business leaders solve their toughest problems. Bain & Co is looking to hire Machine Learning talent to join our Machine Learning Engineering team.
Joining the team you will help develop machine learning pipelines to solve problems of major companies that are clients of Bain & Co. Projects typically last 3-6 months which means that you will see a variety of problems and business cases throughout your career.
Hopefully it’s useful for anyone exploring flow matching for generative modeling. Writing it certainly helped solidify my own understanding.