kshitij_libra

I am building a set of benchmarks and I’d like to publish results.

How can i get peer-reviews, and how does one publish such research independently ? Anyone has experience, willing to guide me on this ?

After using code LLMs' for a bit, I wonder whether there is significant productivity gain in using them. I found that it's limited for me.

Main reason: a) to find the right code, I have to keep sampling it, and b) it doesn't seem to be able to solve larger / more complex problems that I actually find more need for.

I found some interesting research on combining planning-algorithms for complex problems, and some ideas on guiding the LLM's decoding process towards correctness by optimizing it via reward functions and reducing the search space. I've detailed and summarised the main points in the post below.

Questions:

1) Do you find code LLM's really useful? Please share some stories / examples where they help vs they didn't. I'm trying to form a better understanding of their usage

2) Any other research ideas being pursued in this field ? / what are you trying ?

Full post with details here: https://kshitij-banerjee.github.io/2024/04/30/can-llms-produce-better-code/

I keep thinking about this, and it’s a fun game.

AI, Robotics, EVs & crypto - There is a lot going on. If you got 1M$ to be invested now, in one stock/asset, and could only take it out after 10 years - where would you put it? And why ?

I know that figma uses Wasm (from blog), but I am new to WASM, and there seem to be quite a few options (and languages) to choose from.

If one was to build an OSS Figma, what are the latest libraries and toolkits one would use?

I have an idea that I think is a great fix to a gaping gap that exists in the market today. I also have a full time job as a developer .

1 - I want to devote more time to my own startup which is eventually what I want to do.

2 - there are interesting problems at work that I keep running across and can’t help but get involved and solve. My team is also quite fun, so it’s addictive at times to build things together.

3 - I can’t seem to get off hacker news and keep reading random pieces of knowledge on a regular basis .

I am doing way more 2 and 3 than I do 1, probably because 2 and 3 give some immidiete satisfaction whereas I’m a lone wolf on 1. But I know that 1 is more important.

How do you keep being focussed and make more time for 1. How does one prevent himself from falling prey to these other distractions ?