MLL Survey of Vector Databases 2023 Survey of Vector Database Management Systems; a great and comprehensive overview of the recent advancements and challenges in vector database management systems, essential for managing unstructured data in applications like large language models: This paper discusses the unique challenges of vector data management, including issues related to semantic similarity, vector size, and indexing difficulties. The survey delves into various techniques for query processing, storage, and indexing, including vector compression and novel query optimization strategies. It categorizes VDBMSs into native and extended systems, highlighting their specific characteristics. The paper also identifies unresolved issues like selecting appropriate similarity scores and designing efficient hybrid operators. This survey is crucial for machine learning practitioners, offering insights into managing large-scale unstructured data in modern applications. |
|
|---|
|
Microsoft & LF on Recommenders This github repo provides a fantastic compendium of best practices on Recommender Systems, initially maintained by Microsoft and now under the Linux Foundation: This comprehensive toolkit for building recommendation systems offers resources across wide range of classical and deep learning algorithms, detailed Jupyter notebook examples, and utilities for tasks like data preparation, model evaluation, and operationalization. This resource is ideal for production machine learning practitioners, providing robust tools and best practices for developing and deploying recommendation systems in various environments, including extensive documentation and community support for both small-scale and large-scale applications. |
|
|---|
|
Stanford on LLM Hallucination Stopping the hallucination of Large Language Models through few-shot grounding on Wikipedia (97% vs 50% OpenAI): An insightful paper which reinforces the potential of RAG introducing WikiChat, a novel LLM grounded in Wikipedia, designed to address the issue of misinformation in large language model chatbots. By combining LLM-generated responses with Wikipedia-sourced information, WikiChat significantly reduces hallucinations, maintaining high conversational quality and low latency. It achieves 97.3% factual accuracy in simulated conversations, outperforming existing models, especially in handling recent and less popular topics. The system, distilled from GPT-4 to a 7B-parameter LLaMA model, demonstrates the feasibility of creating more reliable, engaging, and efficient chatbots for open-domain conversations. The paper also emphasizes ethical considerations in AI development, including user privacy and fair compensation in crowdsourcing. |
|
|---|
|
DeepMind vs Doctors Turing Test Google DeepMind's latest turing test against Human Doctors (Top-10 Accuracy 59% vs 34%): An insightful research initiative from Google DeepMind based on fine-tuned large language models showing significant promise in medical diagnostics, outperforming human doctors (within specific test cases) in creating differential diagnoses. Their AI system "Articulate Medical Intelligence Explorer (AMIE)" was tested in complex case reports, and included the correct diagnosis in its top 10 list 59% of the time, surpassing board-certified physicians who achieved 34%. Additionally, when used as an interactive assistant, the AI enhanced physicians' diagnostic accuracy from 36% to 52%. Despite its potential, there is an emphasize needed for further real-world testing and consideration of safety and fairness before deploying such AI in healthcare settings. |
|
|---|
|
AI in RecSys Ranking How to reduce cost of ranking by knowledge distillation in Recommender Systems: This article from RecsysML focuses on optimizing machine learning ranking systems by introducing an "early ranker" to reduce computational costs. This early ranker, which uses only about 5% of the compute cost per item compared to the final ranker, is trained through "Knowledge Distillation" to closely mimic the final ranker's decisions, rather than just learning from user actions. This approach addresses alignment issues between the early and final rankers, ensuring more effective and cost-efficient filtering of candidates for final ranking. The article suggests two methods for knowledge distillation and refers to further in-depth resources for interested practitioners. |
|
|---|
|
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 MLOps conferences:
In case you missed our key talks:
|
|
|---|
| | |
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! |
|
|---|
| | |
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!
|
|
|---|
| | |
| | | | The Institute for Ethical AI & Machine Learning is a European research centre that carries out world-class research into responsible machine learning. | | | | |
|
|
|---|
|
|
This email was sent to You received this email because you are registered with The Institute for Ethical AI & Machine Learning's newsletter "The Machine Learning Engineer"
|
| | | | |
|
|
|---|
|
© 2023 The Institute for Ethical AI & Machine Learning |
|
|---|
|
|
|
The Institute for Ethical AI & Machine Learning
