Feeling inspired to write your first TDS post? We’re always open to contributions from new authors.

The articles we feature on our Deep Dives page include detailed walkthroughs of cutting-edge research, explainers on mathematical concepts, and patient tutorials on building and deploying LLM-based tools. Collectively, they represent some of our most thoughtful, in-depth stories.

This week, we invite our community to take a step back from the go-go-go rhythm of daily life and carve out some time to explore a selection of recent Deep Dives—all of which offer nuanced takes on key data science and machine learning topics.

Are you in the mood for tinkering with some code? Would you rather reflect on some of the Big Questions shaping debates around AI? Either way, we’ve got you covered: the lineup we put together in this special edition of the Variable covers a lot of ground, and offers multiple entryways into complex (and fascinating) conversations. Choose your own adventure!

• Gen-AI Safety Landscape: A Guide to the Mitigation Stack for Text-to-Image Models
  “Given the potential risks tied to image generation and inpainting capabilities, it is necessary to establish a robust safety mitigation stack across different stages of the model’s lifecycle.” Trupti Bavalatti unpacks the different approaches currently available to address the inherent risks in generative-AI image tools.
• A Gentle Introduction to the DCIN for Decentralized Inference
  How does a decentralized collaborative intelligence network work? Marcello Politi presents the work he and his team have focused on in recent months: a network of nodes that share computational power to execute inference on open source models, “where computation is distributed dynamically and efficiently, while also maintaining a high level of security and rewarding users for sharing their computation.”
• Les Misérables Social Network Analysis Using Marimo Notebooks and the NetworkX Python Library
  Network analysis has numerous use cases in day-to-day data science workflows—and it can also help us detect patterns and relationships in works of art. Case in point: Maria Mouschoutzi, PhD’s fascinating project, which relies on the NetworkX library and Marimo notebooks to study the intricate social landscape represented in Victor Hugo’s Les Misérables. Whether you’re team Valjean or team Javert—or just into learning about new data science tools—you should add it to your reading list.

• Let There Be Light! Diffusion Models and the Future of Relighting
  Diffusion models made a splashy entrance a few years ago, and researchers have devoted a lot of time and energy into optimizing their performance ever since. Pulkit Gera offers a comprehensive review of recent work on one key aspect for these models: relighting, “the task of rendering a scene under a specified target lighting condition, given an input scene.”
• Lessons in Decision Making from the Monty Hall Problem
  It’s always rewarding to approach a well-known topic from a fresh, thought-provoking angle. That’s exactly what you’ll find in Eyal Kazin’s accessible and comprehensive primer on the the classic Monty Hall problem, looking at it from three distinct perspectives and digging into its underlying math and real-world applications.
• A Critical Look at AI Image Generation
  “The point I want to make is that these are not free of influence from culture and society — whether that influence is good or bad.” Stephanie Kirmer reflects on image-generation models holistically, taking into account the limitations of their aesthetics and the potential biases they reflect.
• Paper Walkthrough: Attention Is All You Need
  It’s likely that no recent ML paper has generated as many guides, explainers, and tutorials as Vaswani et al.’s 2017 landmark contribution. Why share another one, you might ask? Read Muhammad Ardi’s terrific deep dive and we suspect you’ll see the point: it does an excellent job unpacking the Transformer’s key components and balancing theory with hands-on implementation.
• SQL and Data Modelling in Action: A Deep Dive into Data Lakehouses
  For anyone taking their first steps working with databases, complex data architectures, and/or SQL, we highly recommend Sarah Lea’s beginner-friendly—but detailed and meticulous—primer, which tackles the basics of SQL and data modeling for cloud applications.

Thank you for supporting the work of our authors! As we mentioned above, we love publishing articles from new authors, so if you’ve recently written an interesting project walkthrough, tutorial, or theoretical reflection on any of our core topics, don’t hesitate to share it with us.

Until the next Variable,

TDS Team

Network Analysis, Diffusion Models, Data Lakehouses, and More: Our Best Recent Deep Dives was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.

​Feeling inspired to write your first TDS post? We’re always open to contributions from new authors.The articles we feature on our Deep Dives page include detailed walkthroughs of cutting-edge research, explainers on mathematical concepts, and patient tutorials on building and deploying LLM-based tools. Collectively, they represent some of our most thoughtful, in-depth stories.This week, we invite our community to take a step back from the go-go-go rhythm of daily life and carve out some time to explore a selection of recent Deep Dives—all of which offer nuanced takes on key data science and machine learning topics.Are you in the mood for tinkering with some code? Would you rather reflect on some of the Big Questions shaping debates around AI? Either way, we’ve got you covered: the lineup we put together in this special edition of the Variable covers a lot of ground, and offers multiple entryways into complex (and fascinating) conversations. Choose your own adventure!Gen-AI Safety Landscape: A Guide to the Mitigation Stack for Text-to-Image Models“Given the potential risks tied to image generation and inpainting capabilities, it is necessary to establish a robust safety mitigation stack across different stages of the model’s lifecycle.” Trupti Bavalatti unpacks the different approaches currently available to address the inherent risks in generative-AI image tools.A Gentle Introduction to the DCIN for Decentralized InferenceHow does a decentralized collaborative intelligence network work? Marcello Politi presents the work he and his team have focused on in recent months: a network of nodes that share computational power to execute inference on open source models, “where computation is distributed dynamically and efficiently, while also maintaining a high level of security and rewarding users for sharing their computation.”Les Misérables Social Network Analysis Using Marimo Notebooks and the NetworkX Python LibraryNetwork analysis has numerous use cases in day-to-day data science workflows—and it can also help us detect patterns and relationships in works of art. Case in point: Maria Mouschoutzi, PhD’s fascinating project, which relies on the NetworkX library and Marimo notebooks to study the intricate social landscape represented in Victor Hugo’s Les Misérables. Whether you’re team Valjean or team Javert—or just into learning about new data science tools—you should add it to your reading list.Photo by boris misevic on UnsplashLet There Be Light! Diffusion Models and the Future of RelightingDiffusion models made a splashy entrance a few years ago, and researchers have devoted a lot of time and energy into optimizing their performance ever since. Pulkit Gera offers a comprehensive review of recent work on one key aspect for these models: relighting, “the task of rendering a scene under a specified target lighting condition, given an input scene.”Lessons in Decision Making from the Monty Hall ProblemIt’s always rewarding to approach a well-known topic from a fresh, thought-provoking angle. That’s exactly what you’ll find in Eyal Kazin’s accessible and comprehensive primer on the the classic Monty Hall problem, looking at it from three distinct perspectives and digging into its underlying math and real-world applications.A Critical Look at AI Image Generation“The point I want to make is that these are not free of influence from culture and society — whether that influence is good or bad.” Stephanie Kirmer reflects on image-generation models holistically, taking into account the limitations of their aesthetics and the potential biases they reflect.Paper Walkthrough: Attention Is All You NeedIt’s likely that no recent ML paper has generated as many guides, explainers, and tutorials as Vaswani et al.’s 2017 landmark contribution. Why share another one, you might ask? Read Muhammad Ardi’s terrific deep dive and we suspect you’ll see the point: it does an excellent job unpacking the Transformer’s key components and balancing theory with hands-on implementation.SQL and Data Modelling in Action: A Deep Dive into Data LakehousesFor anyone taking their first steps working with databases, complex data architectures, and/or SQL, we highly recommend Sarah Lea’s beginner-friendly—but detailed and meticulous—primer, which tackles the basics of SQL and data modeling for cloud applications.Thank you for supporting the work of our authors! As we mentioned above, we love publishing articles from new authors, so if you’ve recently written an interesting project walkthrough, tutorial, or theoretical reflection on any of our core topics, don’t hesitate to share it with us.Until the next Variable,TDS TeamNetwork Analysis, Diffusion Models, Data Lakehouses, and More: Our Best Recent Deep Dives was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.  tds-features, deep-dives, the-variable, data-science, machine-learning Towards Data Science – MediumRead More

Add to favorites