astrobit @astrobit3

Joined March 2020

Tweets

4K
Followers

109
Following

2K
Likes

24K

Akshay 🚀 @akshay_pachaar

2 years ago

My lecture at MIT!✨ From Physics to Linear Algebra & Machine learning, I have learned a lot from MIT! Yesterday, I had the honour of delivering a guest lecture on The state of AI Engineering, exploring: - Prompt Engineering - Retrieval Augmented Generation. - Fine-Tuning…

105 369 3K 837K 2K

Download Image

Isabel Castro @IsabelcastroG

2 days ago

La solució no és inundar el país amb mà d’obra importada de baixa qualificació. És innovar en tecnologia, de manera que una part de les feines que ara es fan a mà (i malament) les facin màquines. És saber quin model productiu i de país vols.

El Club del Arte 🎨📷📚🖼🕍🎼 @Arteymas_

3 days ago

457 4K 15K 1.3M 1K

Download Video

46 357 848 16K 6

🎸 Rock History 🎸 @historyrock_

19 hours ago

Have you seen Guns N' Roses live?

595 200 3K 117K 136

Download Video

Xavier Climent @ivaxsirc

3 days ago

Estàs cansat d'escoltar sempre la mateixa cantarella? Passa't al heavy

78 258 941 29K 13

Download Image

Sr.Liberal @SrLiberal

3 days ago

Kurt Caz en Roma: “¿Habéis visto cómo nos odian? Vienen a nuestros países pidiendo derechos, nos exigen, la ley les protege y nadie hace nada. Estos son los ingenieros y doctores que están entrando en vuestros países.”

100 2K 5K 138K 210

Download Video

Akshay 🚀 @akshay_pachaar

5 days ago

Today, we are covering the 4 stages of building LLMs from scratch to make them applicable for real-world use cases. We'll cover: - Pre-training - Instruction fine-tuning - Preference fine-tuning - Reasoning fine-tuning The visual summarizes these techniques. Let's dive in!

2 30 245 14K 405

Download Gif

VIEH Group @viehgroup

6 days ago

Find the origin servers of websites protected by Cloudflare, Sucuri, or Incapsula with a misconfigured DNS. ⚔️ - github.com/MrH0wl/Cloudma… Credit: @0x0SojalSec #infosec #bugbountytips #Cybersecurity

1 107 475 21K 450

Download Image

CR1337 @cryptonator1337

7 days ago

If you want to help the Monero network, here is how to mine Monero (XMR) via Gupax: Unlike Bitcoin, which relies heavily on specialized hardware (ASICs), Monero uses the RandomX algorithm, designed for regular CPUs. You can start mining XMR with any decent computer (AMD Ryzen;…

13 28 145 6K 88

Download Image

Sumanth @Sumanth_077

6 days ago

Train multi-step LLM agents with just a few lines of code! Agent Reinforcement Trainer (ART) let's you train LLM agents using reinforcement learning. It makes it easier to build multi-step, reliable agents for real-world tasks using GRPO. 100% Open Source

14 107 506 31K 530

Download Image

ransomNews @ransomnews

7 days ago

⚠️ Trojan in disguise: PDF editor backdoor The “AppSuite PDF Editor” poses as a legit tool but is a backdoor: most of its code supports malware functions, including C2 comms, scheduled tasks, and AES-encrypted data exfiltration. Persistence is full, cleanup needs system repave.…

1 127 617 28K 325

Download Image

𝕄𝕣ℂ𝕣𝕪𝕡 ㉿ priv/acc @MrCrypPrivacy

6 days ago

🚨 The NSA, FBI, IRS-CI, FinCEN, OFAC, Secret Service, Europol, and agencies like ICE don't just monitor cryptocurrencies: they actively use them to track, infiltrate, and dismantle illicit operations 🔍 Tools like Chainalysis Reactor, TRM Labs, Elliptic, MetaSleuth, Cognyte's…

3 24 109 5K 70

Download Image

Víctor Riverola i Morera @riverolavic

6 days ago

Podeu fer un gran RT siusplau?. Moltes gràcies. MANIFEST CONTRA EL “DISTRICTE 11” DE BARCELONA El poble de Barcelona ha estat coneixedor aquests darrers dies de la creació del projecte “Districte 11” que durà el nom de Palestina i Gaza i tindrà un pressupost d’1 milió d’euros.…

69 2K 2K 31K 22

Eyal Yakoby @EYakoby

a week ago

At the University of Amsterdam, Jihadists beat students with sticks. Europe is committing suicide.

8K 25K 91K 6.8M 7K

Download Video

Akshay 🚀 @akshay_pachaar

a week ago

New embedding model cuts vector DB costs by ~200x. It creates embeddings with full document context, unlike traditional models that generate them independently for each chunk. It also outperforms OpenAI and Cohere models. Here's a visual explanation:

Akshay 🚀 @akshay_pachaar

a week ago

23 118 1K 323K 3K

15 69 442 52K 462

Download Gif

elvis @omarsar0

a week ago

RAG is not dead! However, we are in an interesting phase of exploring unique ways to index and retrieve information. This vectorless RAG framework uses a tree structure index in place of vectors. Reasoning models will enable methods that mimic human-like search. Early days!

30 160 1K 91K 1K

Download Image

Unmanned Systems Technology @UnmannedSystem

a week ago

Here's a round-up of the five most read articles on UST this week, featuring new technologies, trials, and industry successes. Read more: hubs.la/Q03G8nm10 #unmanned #aviation #innovation

0 7 17 1K 1

Akshay 🚀 @akshay_pachaar

a week ago

A new embedding model cuts vector DB costs by ~200x. It also outperforms OpenAI and Cohere models. Let's understand how you can use it in LLM apps (with code):

23 118 1K 323K 3K

Sumanth @Sumanth_077

a week ago

Build AI Agents with scalable long-term memory! Mem0 is a fully open source memory layer that provides AI agents with persistent, contextual memory. 100% Open Source

11 90 534 34K 579

Download Image

Jafar Najafov @JafarNajafov

a week ago

🚨 Anthropic just gave away their internal prompt engineering secrets for free. This guide contains the exact techniques their team uses to get perfect AI responses. Here’s what’s inside: