went to college today for an event after so long and somehow ended up getting compliments on my long hair from a group of junior girls 😭
didn't expect that at all
Vaibhav retweeted
I really had more to say but it went too long than expected 🥲👇
And one thing I always say is come out of ur comfort bubble!!!
likitha3126.medium.com/commu…
brooo this is so cool 😭
got raichu and flareon
im keeping both bags 😭
love this @letsblinkit
Vaibhav retweeted
My roadmap goes from Python & data analysis to ML, deep learning, Transformers, LLMs, RAG, and deployment.
The goal isn't to master everything in 70 days. It's to learn consistently, build things, and document the process.
🌐 learnaiwithme.netlify.app/
like basically rn i'm back to my polymath side again
so rn i'm learning ml(progress is crazy like i'm catching up pretty fast 😎), working on 2 side projects,will start dsa and leetcode again 😅, i've also restarted watching anime( don't get much time but like 1-2 eps whenever i can), then after a year i've started playing chess again 🥲,then there's maths no one can escape it i'm revisiting topics and polishing them,github i don't even need to say anything about it. like over the past 2 months a lot of bad stuff happened and i'm genuinely grateful to my close circle who supported me during that time and also pushed me forward some people taught me valuable life lessons😼 also i learnt so much the most imp thing i realised is that i never valued myself 🫠even when i should have because i never took myself seriously and it backfired but not anymore 😎. also i have a lot of thoughts and off-topic things i'd like to share so i'm thinking of publishing them on medium (next article will be a lil personal like about self-respect and my own experience 😇) mostly non-technical stuff and for technical articles ig X is fine and then i'll organise them on github under a repo
well i said all this ig because when you start embracing yourself, respecting yourself,giving yourself a second chance and then working hard and seeing even small results, things start to work out 🗿
and my circle has gotten pretty small but that's what i want, genuine people and i'm grateful to have you guys around. most of the people in my circle are very dedicated to their goals i'm wishing everyone all the best and i genuinely believe they will succeed in their goals and also just don't make harsh or life-changing decisions in one night. it's best to just go and get some sleep at that moment trust me things aren't always as big as they seem at that moment. at the end of the day things do get better it takes time to get back to normal but it does happen😊 don't let certain incidents or people ruin things for you 😇 and also don't misinterpret this as a comeback or any of that bs 😭🙏 it's just a moment i took to express my thoughts and share what's going on in my life currently and i've been inactive for a while now but i'll try to engage more 😁
guess that's enough for a tweet just wanted to share all this anyways that's all from my side for now 😁 thanks to everyone who read till the end ❤️ take care of yourself and keep doing what you love. wishing you all the best with whatever you're working towards 😊
Vaibhav retweeted
Hello. If you are into-
Accounting
Understanding economics.
Learning about business deeply.
Passionate about writing and reading.
Let's connect :)
watched an anime movie today 😁 been pushing myself a lot lately, so this little pause felt nice sometimes a break is part of the journey too
Hey @grok Based on my tweets, which movie character am I ???
wrote something a little different tonight.
not every fight is worth winning.
medium.com/@devvaibhav37/not…
haven't seen anyone on my tl talking about NCM
yet so let me have the honour.
next-concept modeling (NCM) takes a different approach to text generation.
unlike gpt's token-by-token generation, ncm works coarse-to-fine, starting with rough representations and gradually adding details while predicting multiple codes in parallel.
the interesting part? researchers reported 21x faster generation on lm1b.
but a small clarification:
- compared against a diffusion language model baseline at a matched parameter scale, not gpt
- still early research, not a proven replacement for existing llms
- results depend on the task and baseline
paper: arxiv.org/abs/2610.14162
Hi @Support, my account @devvaibhav37 keeps getting flagged for automated activity whenever I like posts.
I completed ID verification today, which resolved the issue for about an hour, but the restriction returned.
Posting and replying work normally. Could you please review this recurring issue?
i don't have any problem posting the unblurred version but it'll reveal my face since it's a real photo converted into anime style 😭
and don't ask about the right side that's a different story 🫣
