Vivek Bhaskar Sathe

global village dweller, proud Indian, proud Singaporean, proud vegetarian, humble human

[माझा आजचा विचार @aajchavichar]

आपटे वाड्यातली पळी — जे मला मिळाले, आणि जे पुढे गेले https://t.co/IXH7VDyvic
#सत्यनारायण_पूजा #satyanarayan_pooja
#मा_आ_वि #majhaaajchavichar

[Global Statistics @Globalstats11]

@PicturesFoIder For anyone wondering how
https://t.co/2OYFZ2NZ5R

[Interesting STEM @InterestingSTEM]

8 forms of intelligence

[Quant Science @quantscience_]

Free Python code from the book:

[Alex Xu @alexxubyte]

Netflix's Tech Stack.

[Massimo @Rainmaker1973]

Animation vs geometry

[🎞️ Alan Becker]
https://t.co/gESZ29UD57

[Abacus.AI @abacusai]

If you’re looking to learn more about prompt engineering, here’s a good survey paper that outlines the different types of prompt engineering (link provided in alt)

Here are some of the strategies discussed:
•Chain of Thought (CoT)
•Chain of Cause (CoC)
•Program-Aided Language Models (PAL)
•Tree of Thoughts (ToT)
•Least-to-Most
•Self-Consistency
•Analogical Reasoning

[Python Coding @clcoding]

Movie Information using Python

[English Grammar tips @knowiiedges]

Words instead of "VERY"

[Bindu Reddy @bindureddy]

Introducing Matrix Agents

Perform complex analysis by creating a matrix agent. You can perform 100s of LLM operations at scale.

Here is what you can do with matrix agents

- portfolio analysis
- stock analysis
- customer and leads analysis
- decision science

This is the first release and we have tons of ideas to improve this class of Enterprise AI agents

[Akshay 🚀 @akshay_pachaar]

This should be impossible!

In just 3 lines of code, this open-source library can clean any ML dataset!

You can:
- flag outliers
- find label errors
- identify near duplicates
- perform active-learning
- find out of distribution samples
- And more ...

Introducing Cleanlab 🚀

A Confident Learning based Python library that works with:

- Any data ( text, image, tabular, audio, etc. )
- Any ML tasks ( classification, tagging, entity recognition, prompting LLMs & more ... )

Check this out👇

You must be wondering how it works! 💭

Cleanlab is developed by folks at MIT, and it's powered by a novel algorithm called Confident Learning!

Below, I have provided a simplified explanation of the same!

Check this out👇

Cleanlab also has a no-code studio that let's you clean any data & train robust models in just a few clicks!

I have shared all the relevant links in the next tweet!

That's a wrap, thanks for reading!

Find me → @akshay_pachaar ✔️
For more content like this!

Cheers! 🥂

[Santiago @svpino]

Lots of people tell me that Cursor + Claude is the best dev setup right now.

I stopped using Cursor a while back. It was too overwhelming.

Trying it again. Looking impressive so far.

[Sumanth @Sumanth_077]

Microsoft launched the best course on Generative AI!

The free 18 lesson course is available on Github and will teach you everything you need to know to start building Generative AI applications.

[Bindu Reddy @bindureddy]

Today, AI can automate 90% of

- customer support
- procurement
- para legal work
- tech and IT support
- HR
- data analyst
- copywriting

We don't need super intelligence; we just need to wire everything up and that's the hard part.

The good news is that it is slowly happening

[Brandon Scott @NaturalMVMT]

Your body moves poorly because you have imbalances.

Structural balance is key to improving not only how you move, but how you age as well.

But what is structural balance??
🧵🧵🧵

[Csaba Kissi @csaba_kissi]

Git essentials

[Bindu Reddy @bindureddy]

The LLM and AGI hype is so over the top that even data science teams are forgetting that LLMs cannot do tabular data ML.

You can't build a great recommender, forecaster, or prediction model with an LLM

You need good ole trees, other tabular data algorithms, and an end-to-end platform for building applied AI systems.

[Reed Strength & Mobility @CoachGeoffReed]

Dr. Peter Attia calls exercise “the most powerful longevity drug.”

But what’s the best way to go about it?

This is a no-brainer for me at 47. Kneesovertoes-style training is fun, endlessly challenging, and works well with my busy weekly schedule as a dad with a demanding, full-time job.

Here’s Day 1 from my @atgexercise “Weekend Warrior Zero” program, a workout you can do anywhere with equipment:

[Elon Musk @elonmusk]

Even Neo had to “Accept cookies to continue”

[Shruti @heyshrutimishra]

Build an AI Research Assistant with GPT-4o, Claude or Llama without any coding knowledge.

(step-by-step instructions):