THE INSTITUTE FOR ETHICAL AIALIGNMENT + SAFETY
JOIN / CONTACT

THE ML ENGINEER — WEEKLY NEWSLETTER

The MachineLearning EngineerIssue #326

Join 70k+ AI professionals receiving weekly curatedarticles, tutorials and blog posts onproduction machine learning.

#325
Issue #32616/03/25generative-ai
#327
Issue #326 🤖 - Google releases Gemma 3, Block Diffusion Challenging LLMs, Mistakes in Large Codebases, Probabilistic Artificial Intelligence, The Startup CTO's Handbook + more 🚀

Thank you for being part of over 70,000+ ML professionals and enthusiasts who receive weekly articles & tutorials on Machine Learning & MLOps 🤖 You can join the newsletter for free at https://ethical.institute/mle.html ⭐

If you like the content please support the newsletter by sharing with your friends via ✉️ Email, 🐦 Twitter, 💼 Linkedin and 📕 Facebook!

This week in Machine Learning:

If you’re looking for an interesting career opportunity, I’m hiring for a few roles including Data Science Manager (Forecasting), as well as Principal Product Manager (Forecasting) - check them out and please do share with your network!

Google releases Gemma 3

Google released Gemma 3 open source - although it’s not yet at DeepSeek-R1 performance-level, the model can run on a single GPU which is impressive on the efficiency side: Google’s Gemma 3 was released in sizes from 1B to 27B and was built to run on a single GPU, and delivers cutting-edge text, image, and code capabilities. It is great to see open weight models that bring advanced multimodal reasoning, and with a hefty 128k-token context window as well as support for over 35 languages (pretrained for 140). It is also interesting to see how responsible AI fits into these releases in practice, Google showcased safety protocols for image safety as well as different testing mechanisms - this space is evolving on a weekly basis.

Block Diffusion Challenging LLMs

Block diffusion is the new kid on the block! It is quite exciting to see that LLMs may finally have reached their parallelization moment: The BD3-LMs introduce a block diffusion framework that brings together autoregressive and diffusion techniques to generate variable-length and long-form text up to ten times longer than prior diffusion models. This is quite exciting research as it basically allows for text generation blocks to be created in a rolling-window basis, which addresses quite a lot of shortcomings from both traditional LLM models as well as previous diffusion approaches - namely enabling for arbitrary length, key-value caching and most importantly enabling parallelization. If new models like this pick up in real-world use-cases, it could have quite a significant improvement on performance and potentially even scale.

Mistakes in Large Codebases

For production ML practitioners it is often important to develop best practices particularly when contributing to large codebases - these are the most common mistakes to avoid: Rather than reinventing the wheel it is very important to align with established patterns to avoid unexpected pitfalls even if your own approach appears cleaner. It is important to thoroughly research how similar functionality like authentication or critical service calls are handled, and mimicking those patterns. It is also important to have an understanding of the hot paths, and ensure cautious testing with a focus on critical scenarios, whilst having a careful approach to adding dependencies or removing legacy code. This is a great article that compiles quite a few good pieces of advice, certainly recommended read for anyone that works in team codebases on a recurrent basis.

Probabilistic Artificial Intelligence

Probabilistic Artificial Intelligence: This is a fantastic university-level resource to polish your foundations on probability and take it to the next level with applied ML models that can be used for autonomous decision making. This 400+ page resource is a great resource for any ML practitioner that are looking to build a deep understanding on uncertainty and probability in machine learning systems. It provides practical resources on probabilistic inference methods such as Bayesian linear regression, Gaussian processes, and Bayesian neural networks, and bridges probabilistic modeling with sequential decision-making techniques such as Bayesian optimization, deep-/model-based reinforcement learning and beyond.

The Startup CTO’s Handbook

The Startup CTO’s Handbook is a great and pragmatic guide for tech leaders transitioning from coding to management in fast-paced (startup) environments: This is now becoming more of a classic (and free) resource that brings together practical frameworks for team building, decision-making for leaders, and techniques for managing technical debt with strategies for effective people leadership.

Upcoming MLOps Events

The MLOps ecosystem continues to grow at break-neck speeds, making it ever harder for us as practitioners to stay up to date with relevant developments. A fantsatic way to keep on-top of relevant resources is through the great community and events that the MLOps and Production ML ecosystem offers. This is the reason why we have started curating a list of upcoming events in the space, which are outlined below.

Upcoming conferences where we’re speaking:

Other upcoming MLOps conferences in 2025:

In case you missed our talks:

Open Source MLOps Tools

Check out the fast-growing ecosystem of production ML tools & frameworks at the github repository which has reached over 10,000 ⭐ github stars. We are currently looking for more libraries to add - if you know of any that are not listed, please let us know or feel free to add a PR. Four featured libraries in the GPU acceleration space are outlined below.

  • Kompute - Blazing fast, lightweight and mobile phone-enabled GPU compute framework optimized for advanced data processing usecases.
  • CuPy - An implementation of NumPy-compatible multi-dimensional array on CUDA. CuPy consists of the core multi-dimensional array class, cupy.ndarray, and many functions on it.
  • Jax - Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
  • CuDF - Built based on the Apache Arrow columnar memory format, cuDF is a GPU DataFrame library for loading, joining, aggregating, filtering, and otherwise manipulating data.

If you know of any open source and open community events that are not listed do give us a heads up so we can add them!

OSS: Policy & Guidelines

As AI systems become more prevalent in society, we face bigger and tougher societal challenges. We have seen a large number of resources that aim to takle these challenges in the form of AI Guidelines, Principles, Ethics Frameworks, etc, however there are so many resources it is hard to navigate. Because of this we started an Open Source initiative that aims to map the ecosystem to make it simpler to navigate. You can find multiple principles in the repo - some examples include the following:

  • MLSecOps Top 10 Vulnerabilities - This is an initiative that aims to further the field of machine learning security by identifying the top 10 most common vulnerabiliites in the machine learning lifecycle as well as best practices.
  • AI & Machine Learning 8 principles for Responsible ML - The Institute for Ethical AI & Machine Learning has put together 8 principles for responsible machine learning that are to be adopted by individuals and delivery teams designing, building and operating machine learning systems.
  • An Evaluation of Guidelines - The Ethics of Ethics; A research paper that analyses multiple Ethics principles.
  • ACM’s Code of Ethics and Professional Conduct - This is the code of ethics that has been put together in 1992 by the Association for Computer Machinery and updated in 2018.

If you know of any guidelines that are not in the “Awesome AI Guidelines” list, please do give us a heads up or feel free to add a pull request!

About us

The Institute for Ethical AI & Machine Learning is a European research centre that carries out world-class research into responsible machine learning.

Check out our website

✉️ Email, 🐦 Twitter, 💼 Linkedin