Prem Sankar G (@premsankar) 's Twitter Profile
Prem Sankar G

@premsankar

Technology strategist | AI/ML, microservices evangelist | Speaker | Opensource contributor. Opinions are my own

ID: 9520862

linkhttp://linkedin.com/in/premsankar calendar_today18-10-2007 13:15:36

5,5K Tweet

1,1K Followers

2,2K Following

Bindu Reddy (@bindureddy) 's Twitter Profile Photo

Research Topics in RAG - Essential For Increasing Accuracy of Custom LLM Apps Building a good custom LLM app is non-trivial and typically involves RAG. While naive RAG is easy to implement, it rarely does well. We still need research in the following key areas so it becomes

Research Topics in RAG - Essential For Increasing Accuracy of Custom LLM Apps

Building a good custom LLM app is non-trivial and typically involves RAG. While naive RAG is easy to implement, it rarely does well.

We still need research in the following key areas so it becomes
The Kobeissi Letter (@kobeissiletter) 's Twitter Profile Photo

Current state of the market: 60 minutes is airing a "rare interview" with Fed Chair Powell tonight at 7:00 PM ET. The interview will "ask about the future of interest rates" and "what the Fed might do next." This comes in the midst of the Fed's biggest interest rate hike

Ronny Haraldsvik (@haraldsvik) 's Twitter Profile Photo

Cohere Technologies & Vodafone Group successfully complete #5G field test of Universal Spectrum Multiplier software which can improve #4G and #5G network capacity by up to 50%... PR at cohere-tech.com

<a href="/Cohere_MultiG/">Cohere Technologies</a> &amp; <a href="/VodafoneGroup/">Vodafone Group</a> successfully complete  #5G field test of Universal Spectrum Multiplier software which can improve #4G and #5G network capacity by up to 50%... PR at  cohere-tech.com
Prem Sankar G (@premsankar) 's Twitter Profile Photo

Learn more about how Microsoft Janus enabled Cohere driven Multi-G ecosystem techcommunity.microsoft.com/t5/azure-for-o…

Andrej Karpathy (@karpathy) 's Twitter Profile Photo

Day 24 of llm.c: we now do multi-GPU training, in bfloat16, with flash attention, directly in ~3000 lines of C/CUDA, and it is FAST! 🚀 We're running ~7% faster than PyTorch nightly, with no asterisks, i.e. this baseline includes all modern & standard bells-and-whistles: mixed

Day 24 of llm.c: we now do multi-GPU training, in bfloat16, with flash attention, directly in ~3000 lines of C/CUDA, and it is FAST! 🚀

We're running ~7% faster than PyTorch nightly, with no asterisks, i.e. this baseline includes all modern &amp; standard bells-and-whistles: mixed
Dushyanth Sridhar (@dushyanthsridar) 's Twitter Profile Photo

SUPPORT OF #Astikas SOLICITED! Kindly share this post(er) with your friends & relatives! #Astikas in the US of A, GC Vedic (Priya & Narayanan) is presenting my tour in 2024. I will be delivering 31 discourses (30 in English, 1 in தமிழ்) in 30 days spread across #NewJersey,

SUPPORT OF #Astikas SOLICITED! Kindly share this post(er) with your friends &amp; relatives! #Astikas in the US of A, GC Vedic (Priya &amp; Narayanan) is presenting my tour in 2024. I will be delivering 31 discourses (30 in English, 1 in தமிழ்) in 30 days spread across #NewJersey,
Andrej Karpathy (@karpathy) 's Twitter Profile Photo

SQL injection-like attack on LLMs with special tokens The decision by LLM tokenizers to parse special tokens in the input string (<s>, <|endoftext|>, etc.), while convenient looking, leads to footguns at best and LLM security vulnerabilities at worst, equivalent to SQL injection

SQL injection-like attack on LLMs with special tokens

The decision by LLM tokenizers to parse special tokens in the input string (&lt;s&gt;, &lt;|endoftext|&gt;, etc.), while convenient looking, leads to footguns at best and LLM security vulnerabilities at worst, equivalent to SQL injection
Alex Albert (@alexalbert__) 's Twitter Profile Photo

Introducing the Model Context Protocol (MCP) An open standard we've been working on at Anthropic that solves a core challenge with LLM apps - connecting them to your data. No more building custom integrations for every data source. MCP provides one protocol to connect them all:

Introducing the Model Context Protocol (MCP)

An open standard we've been working on at Anthropic that solves a core challenge with LLM apps - connecting them to your data.

No more building custom integrations for every data source. MCP provides one protocol to connect them all:
elvis (@omarsar0) 's Twitter Profile Photo

Reverse Thinking Makes LLMs Stronger Reasoners Shows that training LLMs to learn "reverse thinking" helps to improve performance in commonsense, math, and logical reasoning tasks. It claims to outperform a standard fine-tuning method trained on 10x more forward reasoning.

Reverse Thinking Makes LLMs Stronger Reasoners

Shows that training LLMs to learn "reverse thinking" helps to improve performance in commonsense, math, and logical reasoning tasks. 

It claims to outperform a standard fine-tuning method trained on 10x more forward reasoning.
elvis (@omarsar0) 's Twitter Profile Photo

DataLab: A Unified Platform for LLM-Powered Business Intelligence Introduces DataLab, a unified BI platform that integrates an LLM-based agent framework with an augmented computational notebook interface. DataLab achieves state-of-the-art performance on various BI tasks across

DataLab: A Unified Platform for LLM-Powered Business Intelligence

Introduces DataLab, a unified BI platform that integrates an LLM-based agent framework with an augmented computational notebook interface.

DataLab achieves state-of-the-art performance on various BI tasks across
Ghibran Vaibodha (@ghibranvaibodha) 's Twitter Profile Photo

🙏✨ A Year-Long Journey—Now Complete ✨🙏 After 1 year of devotion, I’m humbled to share 5 hours and 44 minutes of Sivavakkiyar’s divine verses. Hearing “ஓடி ஓடி / ஓம் நமசிவாய” in temples feels truly blessed. 🌸 This is more than music—it’s a prayer, a tribute to Lord Shiva. I

🙏✨ A Year-Long Journey—Now Complete ✨🙏

After 1 year of devotion, I’m humbled to share 5 hours and 44 minutes of Sivavakkiyar’s divine verses. Hearing “ஓடி ஓடி / ஓம் நமசிவாய” in temples feels truly blessed. 🌸

This is more than music—it’s a prayer, a tribute to Lord Shiva. I
NVIDIA (@nvidia) 's Twitter Profile Photo

Announcing NVIDIA Project DIGITS, a personal AI supercomputer that’s powered by the NVIDIA GB10 Superchip and based on #NVIDIAGraceBlackwell architecture. nvda.ws/4fLyTm3 Preconfigured with the NVIDIA AI software stack, developers, researchers, data scientists and

Announcing NVIDIA Project DIGITS, a personal AI supercomputer that’s powered by the NVIDIA GB10 Superchip and based on #NVIDIAGraceBlackwell architecture. nvda.ws/4fLyTm3

Preconfigured with the NVIDIA AI software stack, developers, researchers, data scientists and
elvis (@omarsar0) 's Twitter Profile Photo

A Survey of Context Engineering 160+ pages covering the most important research around context engineering for LLMs. This is a must-read! Here are my notes:

A Survey of Context Engineering

160+ pages covering the most important research around context engineering for LLMs.

This is a must-read!

Here are my notes: