Ankit Mudgal retweeted
"Never mock a pain you haven't endured. Life has a funny way of humbling people."
Ankit Mudgal retweeted
Introducing Muse Wonder, our new family of frontier models with state of the art performance for personal and professional agentic work.
Ankit Mudgal retweeted
You can skip all the parties, all the conferences, all the press, all the tweets. Build a great product and get users and win.
Ankit Mudgal retweeted
You can literally type a sentence and watch a Transformer process it.
This interactive explainer runs a GPT-style model in the browser and lets you follow the whole path:
tokenization → embeddings → Q/K/V → attention → MLP → logits → next-token probabilities
You can inspect attention maps, intermediate values, and change temperature, top-k and top-p to see how generation changes.
If Transformers still feel like boxes connected by arrows, this is worth opening.
poloclub.github.io/transform…
Ankit Mudgal retweeted
As companies have invested trillions into AI cap ex, driven by the hope of riches, it is not just markets that have been roiled. I look at how the spending has percolated through financial statements distorting book capital, earnings and cash flows. bit.ly/4zrs8Ao
Desperately waiting for @Starlink services in India. In my newly developed locality. The internet providers @reliancejio @airtelindia says will only lay cables when we get more than 50 connections
Ankit Mudgal retweeted
Starlink in India would enable high-speed, affordable Internet connectivity for those who can’t afford current prices or who don’t have a connection at all!
ELON MUSK: “We’d love to be operating Starlink in India. That would be great.”
It’s been nearly four years since Starlink applied to operate in India, and it is still waiting to launch.
The government should grant the remaining approvals as soon as possible so Starlink can bring high-speed internet to India’s remote and underserved communities.
Ankit Mudgal retweeted
Random German lab appears from nowhere and drops SOTA open-weight model
Beats Qwen, Mistral, Nematron models with similar active parameter count
Trained from stratch
Small bird, fast wings, Kolibri is here.
78B parameters. 3.46B active. Up to 1M tokens of context. Built in Europe.
Now the weights are yours. Run it on your own hardware, under Apache 2.0.
Readers added context they thought people might want to know
x.com/kimmonismus/st…
It beats older generation models and already obsolete nemotron 3 super and mistral 4 small(not even 3.5 medium) and it beats Qwen 3.6 35B a3b(which is very small) not current gen Qwen 3.8 27B and flash next. And the benchmarks are selective, they do not demonstrate coding,etc.
Ankit Mudgal retweeted
I think trading is probably the only profession where a long weekend doesn’t feel exciting ,it feels boring.
Ankit Mudgal retweeted
Oh no! Such horrid people
Fake cancer drugs supplied to more than 90 Bengaluru hospitals, pharmacy owner arrested indianexpress.com/article/ci…
