Yes, temples must be freed if government control.

We absolutely should rewrite our history. Truth about mughal rulers (and their cruelty) must be told. Also, time to bring to mainstream about our great kings like shivaji, raja raja chozhan, krishnadevaraya etc.

Twitter india should not block Anand Ranganathan account.

Most of the dynasties below ruled far more geographical area than the mughals, but unfortunately, our history books have talked about every mughal king, while omitting specific kings of these great dynasties (or the greatest of kings like raja raja chozhan will get a single para)

My team is expanding, so opening up one more role for senior researchers in AI/ML. Senior researchers from both Industry and academia with top AI/ML publications such as NeurIPS, IJCAI, COLT, ICML etc. can apply. We solve some of t…lnkd.in/gAuhPwPR lnkd.in/gkXjTUzb

Interesting work from Microsoft on AI: With more than 1,000 shark and ray species in the ocean, how can we identify suspicious fin shipments into Singapore with better accuracy and efficiency? 🔎🦈 In celebration of #WorldOceansDay 🌊 we’re excited to i…lnkd.in/dE6wsfKX

Interesting work - we did a literature survey on recent work in the Transformer space. The result is this paper, thanks to Badri N. Patro for doing bulk of this work. Thanks to elvis for including our paper in his tweet. x.com/omarsar0/statu…

A few of my old colleagues pinged me to know my opinion on the Future of Life (FLI) open letter () to stop training LLMs larger than GPT-4. I am not a signatory of this letter. It may be infeasible to come up such a strong ban -…lnkd.in/e5tVVszZ lnkd.in/eqbwt4PD

A few friends sought my opinion on the FLI letter and asked whether I am a signatory. Here is a quick post on it. linkedin.com/posts/vijaysri…

Interesting work from MSR in collaboration with academia: HuggingGPT, where the authors have created an LLM to perform 4 major tasks: 1.     Plan – understand user’s intent from the task description and break it down into solvable sub tasks via prompt eng…lnkd.in/gN4PDqJk

SpectFormer: Frequency and Attention is what you need in a Vision Transformer abs: arxiv.org/abs/2304.06446 project page: badripatro.github.io/SpectFormers/

Pretty helpful free pdf editor. Edit PDF - Online PDF Editor and Form Filler ilovepdf.com/edit-pdf #ilovepdf via iLovePDF

Interesting work on scaling transformers. Hoping we can combine this with our SpecFormer idea. Badri N. Patro huggingface.co/papers/2305.07… #ArtificialIntelligence #deeplearning #MachineLearning

Am an honest tax payer looking to withdraw money from my PF account to repay housing loan - they rejected my online withdrawal form twice on flimsy grounds. Not a single phone call being picked. Ministry of Labour & Employment, GoI please help. How come there is no escalation path for requests?

Humbles to have guided Badri in this work!

450 year old Original Ramcharitmanas written by Tulsidas ji is well preserved here in Chitrakoot, Uttar Pradesh.

Glad to announce our paper named “SiMBA-TS: Looking Beyond Transformers for Long Term Time-Series Forecasting” which has been accepted in the ICASSP 2025 (lnkd.in/gcyeRiRK) A short summary of the paper: Thanks Badri N. Patro, my co-author lnkd.in/gi65UnSi

AI solving grad level mathematical problems, that too "original" problems, set by a CMU math professor, which were diffcult for his students. Remarkable. hashtag#artificialintelligence hashtag#deeplearning hashtag#neuralnetworks hashtag#LLMs 📷 Po-Shen LohPo-Shen Loh

Resharing our SpectFormer paper, published in WACV 2025: linkedin.com/posts/vijaysri… Recent citations include 3 from AAAI, 2 from NeurIPS, 4 from IEEE Transactions on GoRS, and 2 from IEEE Transactions on Multimedia. SpectFormer paper is available from lnkd.in/gAvQxGsP

Indiscriminate use of AI. No human-in-the-loop verification. Wonder how they tested groundedness of the LLM before using it for such an important task. LLM hallucination is real.