Ikthedar Chowdhury

Your next door mate who happens to be a software developer.

I write about things that move me as a human being.

Building DeclutteredLife.

Fascinated with 3H.

[Ikthedar Chowdhury @IkthedarC]

Whatever your tech stack is.

Learning about System design and software architecture should be the fundamental step of any software dev learning curve.

Gives you the option to watch the whole process from a bird's eye view & you will know where the blocks or safe paths are.

[Fernando @Franc0Fernand0]

If you're a software engineer who want to get better at system design, read these 10 articles: ↓

[Ikthedar Chowdhury @IkthedarC]

@BhavKhatri @SGarks @jimalkhalili I'd be more concerned about the "consciousness" though. Instead of focusing on "intelligence".

[Ikthedar Chowdhury @IkthedarC]

@BhavKhatri @SGarks @jimalkhalili True, but our human brains also develop intelligence by improving at pattern recognition since childhood, right? Otherwise, a child would be born intelligent.

[Ikthedar Chowdhury @IkthedarC]

Dockerfile (instructions on how to build the ap) --> Docker image (with the app inside) --> Run a Container from the image.

And Docker Daemon is the central service that coordinates all these (from building images to managing containers and resources).

[The AI Timeline @TheAITimeline]

🚨This week’s top AI/ML research papers:

- Mixture-of-Transformers
- BitNet a4.8
- LoRA vs Full Fine-tuning: An Illusion of Equivalence
- Mixtures of In-Context Learners
- Emergence of Hidden Capabilities
- DimensionX
- The Surprising Effectiveness of Test-Time Training for Abstract Reasoning
- OpenCoder: The Open Cookbook for Top-Tier Code LLMs
- ReCapture
- Needle Threading
- M3DocRAG
- Controlling Language and Diffusion Models by Transporting Activations
- Why Do We Need Weight Decay in Modern Deep Learning?
- "Give Me BF16 or Give Me Death"? Trade-Offs in LLM Quantization
- Adaptive Caching for Faster Video Generation with Diffusion Transformers
- Constant Acceleration Flow
- Randomized Autoregressive Visual Generation
- Physics in Next-token Prediction
- In-Context LoRA for Diffusion Transformers
- Balancing Pipeline Parallelism with Vocabulary Parallelism
- EoRA: Eigenspace Low-Rank Approximation
- Self-Consistency Preference Optimization
- How Transformers Solve Propositional Logic Problems: A Mechanistic Analysis
- LASER: Attention with Exponential Transformation
- Photon: Federated LLM Pre-Training
- Attacking Vision-Language Computer Agents via Pop-ups
- Hunyuan-Large
- Context Parallelism for Scalable Million-Token Inference
- Stealing User Prompts from Mixture of Experts
- Deep Learning Through A Telescoping Lens: A Simple Model Provides Empirical Insights On Grokking, Gradient Boosting & Beyond
- Does your LLM truly unlearn? An embarrassingly simple approach to recover unlearned knowledge

overview for each + authors' explanations
read this in thread mode for the best experience

[Quant Science @quantscience_]

Financial Statement Analysis with Large Language Models (LLMs)

A 54-page PDF:

[Elon Musk @elonmusk]

Try out the @xAI API!

[Santiago @svpino]

The best way to learn how to work with Large Language Models is to build one yourself.

Learning is not meant to be fun or easy. It is supposed to be hard work. You are supposed to sweat, stumble, fall, stand, and try again.

This book's tagline is "Master the art of engineering large language models from concept to production." ← Those are precisely the skills you can sell anytime, anywhere.

Here is the link: https://t.co/7fP3e1mcSX

Have fun!

[Ikthedar Chowdhury @IkthedarC]

@therjrajesh Resume

[Rajesh Kumar @therjrajesh]

This Resume has an ATS score of more than 90 🤯

This Resume helped many in getting an interview calls from companies like Google, Microsoft, Amazon, and many more. 💼

I have personally used this single-column resume in my job hunting and got amazing results

I am sharing the exact similar editable ATS Friendly Resumes templates!

Just Free Of Cost!!

Simply:

1. Follow(So that I can DM)
2. Like and Repost
3. Comment "Resume" to receive your copies.

[Ikthedar Chowdhury @IkthedarC]

Deleting Adobe PDF Reader, which has been my go-to for a while. Switching to Sumatra PDF is refreshing – lightweight and efficient, unlike Adobe that eats up so much space. Removing Adobe Acrobat has freed up almost 10GB! #sumatrapdf #adobeacrobat

[maharshi @maharshii]

math looks easier in code, i wonder why

[Ikthedar Chowdhury @IkthedarC]

@svpino Hi Santiago, not directly related to this topic, and I don't have much knowledge on this. But the more dimensions we have in vector databases, the more accurate results we can get from them, is that so?

[Deedy @deedydas]

When you extend the analysis to 100 programs, C / Java are nearly 2.5-3x less information dense than Python.

In other words, Python is the most token efficient way for LLMs to program in

[Ikthedar Chowdhury @IkthedarC]

@huge_icons D but I don't know why. Is it just because we are more familiar with it?

[Ikthedar Chowdhury @IkthedarC]

@_HAWKISH @fleur_de_lune__ @MushtaqBilalPhD Do we think, professors won't be needing graduate students to assist them anymore?

[Ikthedar Chowdhury @IkthedarC]

Is building an AI agent to solve a specific problem more efficient than tuning an existing model?

[Ikthedar Chowdhury @IkthedarC]

My current goal is to take certifications and do hands-on projects on:
1. Data pipeline
2. Docker
3. CI/CD

[Rowan Cheung @rowancheung]

Anthropic just announced Computer Use

It allows Claude to control your computer screen based on a prompt and take actions on your behalf

The use cases in agentic coding with automated debugging, customer support, and education are going to be INSANE

[Sebastian Raschka @rasbt]

The Llama 3.2 1B and 3B models are my favorite LLMs -- small but very capable.
If you want to understand how the architectures look like under the hood, I implemented them from scratch (one of the best ways to learn): https://t.co/ODlwRfONOz