Kevin Du

ML AI engineering

[Kevin Du @Kevin_Duu]

Excellent post on FDE culture.

Hire top engineers who are excellent at building software for the field. Let the field shape the product.

That’s almost the opposite of how every company I’ve worked at builds software.

https://t.co/LWUr5BYKZs

[Kevin Du @Kevin_Duu]

A practical ML pattern: use hindsight to create better training signals, then distill them into a model you can actually deploy.

https://t.co/oMtRKb0PHy

[Kevin Du @Kevin_Duu]

@HamelHusain both human and agents love the topic

[Kevin Du @Kevin_Duu]

https://t.co/YImRtruLAj

[Boris Cherny @bcherny]

Something I have been thinking about: in the past, the best engineers I knew spent a lot of time automating their work in various ways. Better vim/emacs automations, writing lint rules to catch repeat code issues, building up a suite of e2e tests so they don't need to smoke test the app manually. These kinds of things were the highest leverage activities an engineer could do, because it multiplied their own output, which in turn meant they could build more things.

I think many of these automations have become even more important now. This is true for a number of reasons.

First, infra and DevX automation speeds you up. And if you are running an army of agents, each of those agents will be sped up also. More automation == more output per unit of time.

Second, moving things to code improves efficiency. Your agent could fix an issue every time it sees that issue happen, but that uses tokens and might miss cases. If Claude instead writes a lint rule, CI step, or routine, that class of issue can be fully automated forever. This is really what people are talking about when they talk about loops -- it's about automating entire types of busywork rather than solving them one off. This isn't a new idea at all. Engineers have been doing this for a long time!

Third and most importantly, automation makes it possible for others to contribute to the codebase more easily. Increasingly what I am seeing is engineers are contributing to codebases on day one because Claude can navigate the codebase for them, and that non-engineers are able to contribute to a codebase as effectively as engineers can. What gets in the way of both of these is domain knowledge that lives in peoples' heads rather than in automation -- the stuff you used to have to learn when ramping up. What has changed thanks to agents is the domain knowledge that can be encoded as infrastructure is no longer limited to what is expressible in lint rules and types and tests; it can now capture nearly all domain knowledge, encoded as code comments and skills and CLAUDE.md rules and memories. If I put up a PR for an iOS codebase I don't know and a code reviewer rejects it because it doesn't use the right framework, or if a designer builds a new feature and it gets rejected because it doesn't follow the right architectural patterns, these are failures of automation.

Every team should be writing the CLAUDE.md's, REVIEW.md's, skills, and docs that enable agents to productively work in their codebase with zero additional context from the prompter. This sounds crazy, and at the same time is a natural extension of the stuff engineers have always done: automate, and encode domain knowledge as infrastructure. As the model gets smarter and as the harness matures, this task becomes easier. In the meantime, it is on every team to look for ways to convert their domain knowledge to infra so that Claude can write code better, so that code review catches issues automatically, and so the next person working on your codebase can contribute more easily.

[Kevin Du @Kevin_Duu]

Great blog on the next iteration of thinking models. By introducing an environment during training, models can learn how to interact with it from day one, unlocking more powerful capabilities. Looking forward to what’s next.

[Junyang Lin @JustinLin610]

https://t.co/mgVhhZVTKg

[Kevin Du @Kevin_Duu]

Good weekend reading while we build in the agentic AI world https://t.co/mTfjgCwaIX

[Andrew Ng @AndrewYNg]

I'm excited to announce Context Hub, an open tool that gives your coding agent the up-to-date API documentation it needs. Install it and prompt your agent to use it to fetch curated docs via a simple CLI. (See image.)

Why this matters: Coding agents often use outdated APIs and hallucinate parameters. For example, when I ask Claude Code to call OpenAI's GPT-5.2, it uses the older chat completions API instead of the newer responses API, even though the newer one has been out for a year. Context Hub solves this.

Context Hub is also designed to get smarter over time. Agents can annotate docs with notes — if your agent discovers a workaround, it can save it and doesn't have to rediscover it next session. Longer term, we're building toward agents sharing what they learn with each other, so the whole community benefits.

Thanks Rohit Prsad and Xin Ye for working with me on this!

npm install -g @aisuite/chub

GitHub: https://t.co/OCkyxXQMCq

[Hōrōshi バガボンド @KatanaLarp]

https://t.co/h9C7YZeCx3

[Raunak @raunakdoesdev]

https://t.co/gQKDSr30sH

[Kevin Du @Kevin_Duu]

@jxnlco @levelsio that will be amazing

[João Moura @joaomdmoura]

https://t.co/VWfQ0pWzcR

[Kevin Du @Kevin_Duu]

How do we keep up with the growing volume of AI agents writing code?
If code review itself has to be delegated, what production gates should we put in place to maintain quality and safety? Great read: https://t.co/ejdAeOH6rV

[@abdimoalim.bsky.social @abdimoalim_]

Fast LLM inference from scratch (by Andrew Chan): https://t.co/Je9aRxhxzW

[Kevin Du @Kevin_Duu]

“A sandbox that can’t access your internal systems is simply a toy.” Nice saying. https://t.co/pKhukJyPVv

[Sakana AI @SakanaAILabs]

We’re excited to introduce Doc-to-LoRA and Text-to-LoRA, two related research exploring how to make LLM customization faster and more accessible.

https://t.co/ApVzVsBuv1

By training a Hypernetwork to generate LoRA adapters on the fly, these methods allow models to instantly internalize new information or adapt to new tasks.

Biological systems naturally rely on two key cognitive abilities: durable long-term memory to store facts, and rapid adaptation to handle new tasks given limited sensory cues. While modern LLMs are highly capable, they still lack this flexibility. Traditionally, adding long-term memory or adapting an LLM to a specific downstream task requires an expensive and time-consuming model update, such as fine-tuning or context distillation, or relies on memory-intensive long prompts.

To bypass these limitations, our work focuses on the concept of cost amortization. We pay the meta-training cost once to train a hypernetwork capable of producing tasks or document specific LoRAs on demand. This turns what used to be a heavy engineering pipeline into a single, inexpensive forward pass. Instead of performing per-task optimization, the hypernetwork meta-learns update rules to instantly modify an LLM given a new task description or a long document.

In our experiments, Text-to-LoRA successfully specializes models to unseen tasks using just a natural language description. Building on this, Doc-to-LoRA is able to internalize factual documents. On a needle-in-a-haystack task, Doc-to-LoRA achieves near-perfect accuracy on instances five times longer than the base model's context window. It can even generalize to transfer visual information from a vision-language model into a text-only LLM, allowing it to classify images purely through internalized weights.

Importantly, both methods run with sub-second latency, enabling rapid experimentation while avoiding the overhead of traditional model updates. This approach is a step towards lowering the technical barriers of model customization, allowing end-users to specialize foundation models via simple text inputs. We have released our code and papers for the community to explore.

Doc-to-LoRA
Paper: https://t.co/87xEEpf0GN
Code: https://t.co/zBfQi2L9LW

Text-to-LoRA
Paper: https://t.co/emLRZ4Vdvo
Code: https://t.co/b9mrdoWWRB

[Andrej Karpathy @karpathy]

Bought a new Mac mini to properly tinker with claws over the weekend. The apple store person told me they are selling like hotcakes and everyone is confused :)

I'm definitely a bit sus'd to run OpenClaw specifically - giving my private data/keys to 400K lines of vibe coded monster that is being actively attacked at scale is not very appealing at all. Already seeing reports of exposed instances, RCE vulnerabilities, supply chain poisoning, malicious or compromised skills in the registry, it feels like a complete wild west and a security nightmare. But I do love the concept and I think that just like LLM agents were a new layer on top of LLMs, Claws are now a new layer on top of LLM agents, taking the orchestration, scheduling, context, tool calls and a kind of persistence to a next level.

Looking around, and given that the high level idea is clear, there are a lot of smaller Claws starting to pop out. For example, on a quick skim NanoClaw looks really interesting in that the core engine is ~4000 lines of code (fits into both my head and that of AI agents, so it feels manageable, auditable, flexible, etc.) and runs everything in containers by default. I also love their approach to configurability - it's not done via config files it's done via skills! For example, /add-telegram instructs your AI agent how to modify the actual code to integrate Telegram. I haven't come across this yet and it slightly blew my mind earlier today as a new, AI-enabled approach to preventing config mess and if-then-else monsters. Basically - the implied new meta is to write the most maximally forkable repo and then have skills that fork it into any desired more exotic configuration. Very cool.

Anyway there are many others - e.g. nanobot, zeroclaw, ironclaw, picoclaw (lol @ prefixes). There are also cloud-hosted alternatives but tbh I don't love these because it feels much harder to tinker with. In particular, local setup allows easy connection to home automation gadgets on the local network. And I don't know, there is something aesthetically pleasing about there being a physical device 'possessed' by a little ghost of a personal digital house elf.

Not 100% sure what my setup ends up looking like just yet but Claws are an awesome, exciting new layer of the AI stack.

[Chris @criccomini]

The 2nd edition of Designing Data-Intensive Applications, by @martinkl and me, is finished and sent to the printers! Ebooks available next week, and print books in 3–4 weeks. Sigh of relief. 😅

(BTW, this is a good opportunity to support your favourite local bookshop!)

[Andrej Karpathy @karpathy]

New art project.
Train and inference GPT in 243 lines of pure, dependency-free Python. This is the *full* algorithmic content of what is needed. Everything else is just for efficiency. I cannot simplify this any further.
https://t.co/HmiRrQugnP

[Kevin Du @Kevin_Duu]

@chipro My job is to monitor business metrics, investigate issues when they arise, find patterns, and build ML models or rules to restore performance. With 95% of data already digitized, the main barrier to automation is whether the effort justifies the automation cost.