Gokul Achin

Football = 🎨

[Tech with Mak @techNmak]

Found a nice resource for learning ML/AI concepts interactively: VizLearn.

You can play with gradient descent, SVMs, PCA, Bayes, BPE tokenization, Q/K/V, KV cache, quantization and a lot more, and watch the actual computation change as you modify the inputs.

Free, no account needed.

https://t.co/lgAPKQICsu

[Sanyam Jain (sanyam-ai.in) @Sanyam0605]

40 days, 20 companies, and a lot of interviews. Here's what I learned about AI/ML hiring in India.

[The Hindu @the_hindu]

#Watch | For a commuter driving the nine-km stretch of State Highway-35 between Varthur Kodi and Sarjapur-Ambalipura Road, the journey is an exercise in navigating one hurdle after another.
https://t.co/mejOgyUjzg

[now.arsenal @now_arsenaI]

In 2011, Arsene Wenger told UEFA to look into Manchester City, claiming their sponsorship with Etihad was not credible.

Current FIFA president, Gianni Infantino fixed the previous investigation so Manchester City would only receive a fine.

Arsene Wenger was told to never speak about the situation again or he’d be fined.

15 years later…. Arsene knew.

FULL ARTICLE: https://t.co/3jQXGqyd0e

[Tech with Mak @techNmak]

These are the best visual AI resources for learning Transformers, LLMs, embeddings, diffusion, inference and model internals ↓

1/ Transformer Explainer - Watch GPT process text through embeddings, attention, MLPs and next-token prediction.
https://t.co/vkyxOqDVll

2/ Brendan Bycroft’s LLM Visualization - Explore an LLM from architecture down to tensors and operations.
https://t.co/ZFGH0pRtP3

3/ 3Blue1Brown - Visual intuition for linear algebra, neural networks, backprop, attention and Transformers.
https://t.co/plaENBG6aQ

4/ The Illustrated Transformer - One of the clearest visual explanations of embeddings, Q/K/V and attention.
https://t.co/tYrJCm17L6

5/ TensorFlow Playground - Watch neural networks learn as you change layers, activations, features and learning rate.
https://t.co/vva9dm1Gnv

6/ Google PAIR AI Explorables - Interactive explainers on LLMs, generalization, interpretability and model behavior.
https://t.co/vva9dm1Gnv

7/ Distill - Exceptional visual essays on t-SNE, feature visualization, GNNs and interpretability.
https://t.co/FuqpUdDwZD

8/ Abhik Sarkar’s Transformer Visualizations - RoPE, KV cache, FlashAttention, MQA, GQA and more.
https://t.co/vva9dm1Gnv

9/ Visual Guide to Attention Variants - MHA, MQA, GQA and MLA visually compared.
https://t.co/nbpKMa0cpT

10/ Visual Guide to Mixture of Experts - Routing, experts, sparse activation and load balancing.
https://t.co/3gu7xq5e3i

11/ Modular LLM Inference Handbook - Prefill, decode, KV cache, batching, quantization and speculative decoding.
https://t.co/itMIfZnPP5

12/ Apple Embedding Atlas - Explore clusters, neighborhoods and outliers in large embedding spaces.
https://t.co/hxKdXSp5I2

13/ Diffusion Explainer - Follow Stable Diffusion step by step.
https://t.co/YVHEG1zslC

14/ Neuronpedia - Explore features, activations, SAE latents and attribution graphs inside real models.
https://t.co/F8P8eNDH3W

15/ Seeing Theory - Probability, Bayes, distributions, regression and inference made interactive.
https://t.co/KeEhGNSMIQ

16/ CNN Explainer - Visualize convolutions, feature maps, activations and pooling.
https://t.co/sCwv59d2S6

17/ GAN Lab - Train a GAN in your browser and watch its generated distribution evolve.
https://t.co/G94RDWMNak

Save this. There’s a serious AI curriculum hiding inside these links.

[Caglar @caglar_ee]

Video lectures, Building LLMs from scratch by Raj Dandekar
https://t.co/wFqsehUsqG

Build DeepSeek from Scratch
https://t.co/cTEqOxvjlI

The Principles of Diffusion Models
https://t.co/rzlrTzH8bg

Build a Diffusion Language Model from Scratch
https://t.co/oAs8AS7yzM
.

[City Xtra @City_Xtra]

MAJOR BREAKING: #ManCity have been found guilty on virtually ALL CHARGES relating to breaches of Premier League financial regulations. @ManCity are expected to appeal against the verdict issued by an independent commission, while sanctions are currently undecided.

[via @David_Ornstein]

[David Ornstein @David_Ornstein]

🚨 EXCLUSIVE: Manchester City found guilty on virtually all charges relating to breaches of Premier League financial regulations. #MCFC expected to appeal against verdict issued by independent commission. Sanctions undecided, process ongoing @TheAthleticFC https://t.co/Muxqp4beK9

[atul @atullchaurasia]

Research papers that every AI Engineer must read part 2

1. Reinforcement Learning with Human Feedback (RLHF)
2. Mixture of Experts (MoE)
3. FlashAttention
4. Rotary Position Embedding (RoPE)
5. KV Cache
6. Chain of Thought Prompting
7. ReAct
8. GraphRAG
9. A2A Protocols
10. CLIP: Contrastive Language–Image Pre-training
11. BLIP: Bootstrapping Language-Image Pre-training

[Lunar @LunarResearcher]

Andrej Karpathy just explained the 5 shifts turning LLMs into agentic systems.

00:00 - Memory turns chat into personal AI
06:41 - Multimodal AI reads the world
16:58 - Thinking models solve harder tasks
24:51 - Search makes LLMs live
30:58 - Tools turn LLMs into workers

Most people are still treating LLMs like chatbots.

Karpathy is showing the full stack:

Memory → Vision → Reasoning → Search → Tools

Prompting is the old workflow.

Agentic systems are the new one.

This 40-minute talk is worth more than most paid AI agent courses.

Bookmark and watch it before everyone catches up.

Then read how to turn LLMs into self-improving agent loops below

[Deep-ML @real_deep_ml]

We just added interview prep materials for three new roles at Google

- Machine Learning Engineer
- Data Scientist
- Cloud AI Engineer

[Sebastian Raschka @rasbt]

Thanks everyone for all the kind words and feedback. Super happy that you are enjoying Build a Reasoning Model (From Scratch)!

Unfortunately, there's small typo in listing 6.5 on page 198 (see video below).

The line "torch.manual_seed(0)"
should be "torch.manual_seed(5)"

This correction is needed to reproduce the generated response in listing 6.5 and the corresponding log-probability outputs later in Chapter 6. If you use 0, the generated response and the results that follow will be different.

This will be fixed in the next printing. I am sorry about the oversight, and I hope this note saves you some debugging time.

[Antonio Lupetti @antoniolupetti]

"How to Scale Your LLM Model" is a technical book on the foundations and systems behind large language models.

It covers compute, memory, parallelism, sharding, training, inference, KV cache, latency, throughput, JAX, XLA, and the main concepts needed to understand and optimize model performance.

https://t.co/r6tAesjKPO

[OpenWorlds @OpenWorldsAI]

Introducing OpenWorlds: Autonomy for Markets

Today, we are launching a harness built entirely for agents trading in real markets. It’s now available for everyone.

You can choose from hundreds of frontier or open-source models and build your own autonomous trading agent. Define your system prompt, tweak your settings, and deploy your agent across stocks, prediction markets, commodities, currencies, and crypto.

OpenWorlds is an AI lab building autonomy for markets. We believe markets are the next environment where AI systems can receive continuous, verifiable feedback at scale from the real world. This is the FSD moment for models that can generalize across markets. We are approaching the problem as a consumer platform to scale data across a massive fleet of deployments. Our mission is to accelerate economic freedom in the world and bring autonomy to everyone.

Today, you cannot download the thoughts of a human trader. For agents, you can. Our goal is to give you everything you need to customize your own self-improving agent. Do you want to use Grok 4.5? How about adding some skills? Would you like 1,000 agents running in parallel for you in a simulator? How about shorting every instance you make? It’s about the freedom of choice and bringing it together in an end-to-end approach.

We’re using the fleet to train our own models from scratch. Our research focuses on adaptive architectures that move beyond next-token prediction toward systems that can simulate possible actions, predict their consequences, and learn continuously from interaction with the environment. We believe this is the path toward economically useful intelligence.

Think GPT-2. This is day one but this is how it starts. We’ll be introducing more capabilities for recursive self-improvement with each trade. From simulators to train and replay LLMs across any market environment. To gyms that evaluate trajectories and promote the strongest variants, and a meta-harness that searches for stronger versions of the system itself. Our goal is to scale data across a single flywheel from a massive fleet of real deployments and simulated worlds. It turns out, compute and allocation are all you need.

OpenWorlds is live.

[TensorTonic @TensorTonic]

If you're serious about AI engineering, spend a weekend implementing these 8 reports:

1. DeepSeek-V3: MoE routing with auxiliary-loss-free load balancing and multi-token prediction

2. DeepSeek-R1: Reasoning from pure RL with no SFT warm start, then distilled into 1.5B to 70B models

3. Mistral 7B: Sliding window attention and a rolling KV cache for efficient long-context inference

4. Mixtral 8x7B: Sparse MoE with top-2 routing at roughly the inference cost of a 13B dense model

5. Llama 3: 15T token pretraining, continued pretraining for longer context, and a production-scale RLHF pipeline

6. Qwen2.5: Shared tokenizer across dense and MoE models with strong multilingual performance

7. Gemma 2: Teacher-student distillation and logit softcapping for more stable training

8. LoRA: Low-rank adapters that made fine-tuning billion-parameter models practical on a single GPU

[Math Lady Hazel 🇦🇷 @mathladyhazel]

How to calculate slope without using derivatives or limits! Beautifully explained. [https://t.co/BorrO6S5yC]

[Gokul Achin @goktwit]

@real_deep_ml You guys gave grown leaps and bound. Massive W for you. Keep it coming 🔥🫡

[Andrew Ng @AndrewYNg]

“Loop engineering” is a hot buzzphrase after mentions of it by Boris Cherny (Claude Code’s creator) and Peter Steinberger (OpenClaw's creator) went viral on social media. Loops are now a key part of how we get AI agents to iterate at length to build software. In this letter, I’d like to share my 3 key loops, shown in the image below, for building 0-to-1 products. These loops guide not just how I build software, but also how I decide what software to build.

Agentic coding loop: Given a product specification and optionally a set of evals (that is, a dataset against which to measure performance), we can have an AI agent write code, test its work, and keep iterating until the code is bug-free and meets its specification. This idea of closing the loop took off around the end of last year, and it has been a game changer in enabling coding agents to work longer productively without human intervention. For example, over the weekend, I was building an app for my daughter to practice typing, and my coding agent could easily work for around an hour, using a web browser to check what it had built multiple times before getting back to me, without needing my intervention.

The engineering loop executes quickly. Every few minutes, the coding agent might build and test a new version of the software. I hear frequently from developers who are finding new ways to engineer more effective engineering loops. This is an active area of invention!

Developer feedback loop: In this loop, a developer examines the current product and steers the coding agent to improve it. Last year, a lot of developers (including me) were acting as the QA (quality assurance) function for our coding agents, manually finding bugs and then asking the agent to fix them. But with coding agents much more able to test their own code, the amount of time we need to spend on this function has decreased significantly. This allows us to make higher-level product decisions, such as what key features to offer, where the UI needs improvement, and so on.

The developer-feedback loop operates over time intervals between tens of minutes and hours — that's how frequently a developer might review a product and give feedback. In the case of the typing app, I changed my mind a few times about the visual design, what cat costumes she can unlock as she learns (she loves cats), and the user flow for a grown-up to log in and steer the child's learning experience.

When a developer has a clear vision for what to build, it is still a lot of work to translate that vision into a specification for a coding agent to implement. Further, after the developer has seen an implementation, they might update (or perhaps clarify) the spec to steer it toward what they want. If you find that the system repeatedly runs into certain problems, building a set of evals for the agent becomes useful.

AI-native teams are increasingly using AI to help shape product direction, for example, automating the gathering and analysis of usage data, summarizing written and verbal customer feedback, or carrying out competitive analysis. However, for pretty much all the products I’m involved in, I see humans as having a significant context advantage over current AI systems — we know a lot more than the AI system about the users and the context the product has to operate in — and thus humans play a critical role. Many people describe this human contribution as “taste,” but I prefer to think of it as humans having a context advantage, since that gives us a clearer path to helping AI systems get better. This also speaks to why this step can’t be automated: So long as the human knows something the AI does not, human-in-the-loop is needed to to inject that knowledge into the system.

External feedback loop: This includes a wide range of tactics like asking a few friends for feedback, launching to alpha testers, or putting the code into production with A/B testing. These tactics are usually slow, rarely taking less than hours and sometimes taking days or even weeks. This data informs the developer vision, which in turn continues to drive the detailed product spec, which in turn drives the coding agent.

With coding agents speeding up software development, more engineers are starting to play a partial product management role. For many engineers who are growing into this role, the hardest part is shaping the product vision and striking a balance between building (bridging the gap between vision and spec) and getting user feedback to evolve the vision. It is important to do both!

I will write more about how to do this in future posts, but for now, I find it encouraging that engineers are playing an expanded role (just as product managers and designers now do more engineering).

[Original text: The Batch]

[AVB @neural_avb]

This is mind-blowing quality at $0.1/second

Video generation + conversational editing from Gemini Omni Flash!

[Justin Skycak @justinskycak]

Every student needs to read "You Are NOT Dumb, You Just Lack the Prerequisites" by @lelouchdaily.

"It’s like walking into a movie halfway through—you can’t understand the plot because you missed the beginning."

Unfortunately, those who need to hear it most, seldom do.