so much how time do you take for this process? are there research papers you wrestle with for weeks?
@deeplearningexplained3 ай бұрын
It really depends on the goal. If I'm trying to understand it deeply enough to mechanistically understand how the methodology work, this can take a few days to a week. If I'm trying to reproduce the paper result to incorporate it within my own research project, I might wrestle with the thing for a full month. However, I don't do this process with all papers. That would be wasteful. I select which paper have high relevancy for myself and my lab by reading lots of them superficially, then I pin point a few that have disproportionate benefits.
@davidlaidbiggestfanАй бұрын
This is phenomenal , can't wait to try this out !
@deeplearningexplainedАй бұрын
Let me know how it goes!
@futureshockpod3 ай бұрын
Hey Yacine, really agree with your points here, they're excellently conveyed. Great new channel I've found!
@deeplearningexplained3 ай бұрын
I’m glad it was useful, welcome to the channel :)
@benxneo3 ай бұрын
could you dive more into how you take notes for papers? Also, Eric Jang has these questions for understanding any ML papers quickly 1. What are the inputs to the function approximator? 2. What are the outputs to the function approximator? 3. What loss supervises the output predictions? What assumptions about the world does this particular objective make? 4. Once trained, what is the model able to generalize to, in regards to input/output pairs it hasn’t seen before? 5. Are the claims in the paper falsifiable? do you think its missing anything?
@deeplearningexplained3 ай бұрын
Yes for sure, I'll be doing a deep technical dive on Segment Anything architecture so I'll take the opportunity to document the process from start to finish. These 5 points are good and cover most of the ground. The one thing I would add before anything else is: "0. How would the model be used in a real-world setting?" If you have a solid answer to that question, you will be able to better understand the method section (1-4) and find critical flaws in the author's reasoning (5).
@mamotivatedАй бұрын
Another great fluid video on an effective process to make progress with research.