user avatar
lovish
Recursive
@louvishh
founding member @recursive_si | phd @ucl and msl @aiatmeta | previously @googleai. mostly random tweets here.
london
lovishmadaan.github.io
Joined July 2021
1,007
Following
1,583
Followers
  • Pinned
    user avatar
    lovish
    Recursive
    @louvishh
    Oct 16, 2025
    🚨 New Paper: The Art of Scaling Reinforcement Learning Compute for LLMs 🚨 We burnt a lot of GPU-hours to provide the community with the first open, large-scale systematic study on RL scaling for LLMs. x.com/Devvrit_Khatri…
    user avatar
    Devvrit
    @Devvrit_Khatri
    Oct 16, 2025
    Wish to build scaling laws for RL but not sure how to scale? Or what scales? Or would RL even scale predictably? We introduce: The Art of Scaling Reinforcement Learning Compute for LLMs
    30K
  • user avatar
    lovish
    Recursive
    @louvishh
    Mar 20, 2022
    exactly 365 days ago, something snapped in my mind, and i finally bought a gym membership. happy then, happier now 🙂
  • user avatar
    lovish
    Recursive
    @louvishh
    Jun 17, 2024
    🚨 New Paper 🚨 Evaluations can have a lot of variance, throwing off model comparisons especially during pre-training. In our latest work, “Quantifying Variance in Evaluation Benchmarks”, we explore this phenomenon in depth. A thread [1/n]
    19K
  • user avatar
    lovish
    Recursive
    @louvishh
    Jul 23, 2024
    405b is out! working on llama 3 has been a truly rewarding experience and i'm super grateful to all my teammates! i'm excited to see how the llama models will be used by the community! p.s. - we wrote a paper and not just a tech report 😛
    15K
  • user avatar
    lovish
    Recursive
    @louvishh
    Oct 13, 2023
    life update ✨: i’ve moved to london and started my phd at @ucl_nlp and @AIatMeta! looking forward to collaborating with folks in research and making new friends. if you’re around in the uk and would like to say hi, please feel free to reach out!
    7.4K
  • user avatar
    lovish
    Recursive
    @louvishh
    Jan 21, 2023
    celebrating iclr acceptance with a swim
    4.7K
  • user avatar
    lovish
    Recursive
    @louvishh
    Mar 20, 2022
    Replying to @louvishh
    just remove sugar, cook in olive oil, bake/grill instead of fry. and of course, you can have cheat days too. and most importantly, don’t worry if there are off days. i had quite a few relapses where i was eating anything i wanted to. just pick yourself up and start again 😁
  • user avatar
    lovish
    Recursive
    @louvishh
    Aug 20, 2022
    another one of my schoolmates is getting married, and here i am, still contemplating if i should apply for a phd.
  • user avatar
    lovish
    Recursive
    @louvishh
    Mar 20, 2022
    Replying to @louvishh
    and diet is also a very important part of this process. i tried a bunch of things - keto/intermittent/calorie deficit etc. i found that you don’t have to follow anything extreme and calorie deficit diets are just as good, and in fact more sustainable for long term.
  • user avatar
    lovish
    Recursive
    @louvishh
    Mar 20, 2022
    Replying to @louvishh
    since a lot of you’ve been asking, for the workouts, i focused on both strength training and cardio. cardio is important for burning calories and weights for building muscle. my average workout routine is 30 mins cardio + 35-45 mins strength training for 5/6 days a week.
  • user avatar
    lovish
    Recursive
    @louvishh
    Mar 4, 2022
    working a full time job and still being handed 500 rupaye ka note forcefully by relatives whenever you meet them has to be the most desi thing ever ffs.
  • user avatar
    lovish
    Recursive
    @louvishh
    Jul 28, 2024
    Replying to @giffmana
    we use our internal evals repository to run all the evals. we did release the inputs/outputs/metrics using our repo for most evaluation tasks (including mmlu) here:
    meta-llama/Llama-3.1-405B-Instruct-evals · Datasets at Hugging Face
    From huggingface.co
    1.3K
  • user avatar
    lovish
    Recursive
    @louvishh
    Apr 29, 2023
    in kigali, rwanda for #ICLR2023! hit me up if you would like to talk about nlp, large language models, and optimization! also stop by our poster: iclr.cc/virtual/2023/p…
    2K
  • user avatar
    lovish
    Recursive
    @louvishh
    Jun 25, 2023
    can never do this in blr. 🚴🏼🚴🏼
    1.1K