Aug 29, 2018 · newsletter

Understanding a generative space

Kate Compton, maker of many interesting things, recently tweeted a diagram from her work-in-progress dissertation zine. The illustration breaks down how we interact with a procedural generator, which can be anything from a photo filter to a character creation tool.

A diagram. The top half shows the expectation space, and the bottom half the algorithmic space. The process is circular. You start with a mental model, make a hypothesis about what a change will do, then encode the change. In the algorithmic space the model is modified by the change, the code reinterprets the model and displays the results of the change. The user then evaluates the artifact and updates their mental model of how the algorithmic space works. The process restarts.

“How we understand a generative space” by Kate Compton. Understanding a generative space requires moving between the digital space (where the code is and the results are displayed) and the mental space (where we evaluate the display and make guesses about how it works).

As app creators, we tend to focus on the parts we can see: the user interface and the code that determines app behavior. Those are the parts that we can directly control. Compton’s diagram is a great reminder that there’s another dimension at work in a user’s experience of an app: their mental model of how the application works. We have much less direct control over a user’s mental model.

The mental model is the user’s “parallel universe” reconstruction of the logic behind an application. As app creators, we try to shepherd the user so their mental model usefully aligns with our own. We employ metaphors, often based on familiar patterns from the physical world or other applications - but there’s always going to be slippage. Part of the fun of designing an application is figuring out how to minimize this slippage, and providing lots of entry points, so that if a user’s mental model gets really misaligned, there are spots within the application where expectations can be reset. As Compton’s diagram shows, one of the best ways to encourage alignment of a mental model and the model itself is to allow the user to make changes and quickly see the results. Enabling that sort of iterative play taps into the human impulse to poke things in order to figure out how they work.

Understanding the interplay between a user’s mental model and an application will be especially important for creating machine learning tools that open up new creative possibilities. Because of their complexity, it may not be feasible for users to develop a comprehensive mental model of how an ML model works. It will then be up to application creators to guide users towards a simplified or alternate mental model that helps them collaborate fruitfully with the application.

Read more

Aug 29, 2018 · newsletter
Aug 29, 2018 · newsletter

Latest posts

Jun 22, 2020 · post

How to Explain HuggingFace BERT for Question Answering NLP Models with TF 2.0

by Victor · Figure 1: In this sample, a BERTbase model gets the answer correct (Achaemenid Persia). Model gradients show that the token “subordinate ..” is impactful in the selection of an answer to the question “Macedonia was under the rule of which country?". This makes sense .. good for BERTbase. Recently, our team at Fast Forward Labs have been exploring state of the art models for Question Answering and have used the rather excellent HuggingFace transformers library. more
Jun 16, 2020 · notebook

Evaluating QA: Metrics, Predictions, and the Null Response →

by Melanie · A deep dive into computing QA predictions and when to tell BERT to zip it! In our last post, Building a QA System with BERT on Wikipedia, we used the HuggingFace framework to train BERT on the SQuAD2.0 dataset and built a simple QA system on top of the Wikipedia search engine. This time, we’ll look at how to assess the quality of a BERT-like model for Question Answering.
May 19, 2020 · notebook

Building a QA System with BERT on Wikipedia →

by Melanie · So you’ve decided to build a QA system. You want to start with something simple and general so you plan to make it open domain using Wikipedia as a corpus for answering questions. You want to use the best NLP that your compute resources allow (you’re lucky enough to have access to a GPU) so you’re going to focus on the big, flashy Transformer models that are all the rage these days.
Apr 28, 2020 · notebook

Intro to Automated Question Answering →

by Melanie · Welcome to the first edition of the Cloudera Fast Forward blog on Natural Language Processing for Question Answering! Throughout this series, we’ll build a Question Answering (QA) system with off-the-shelf algorithms and libraries and blog about our process and what we find along the way. We hope to wind up with a beginning-to-end documentary that provides:
Apr 1, 2020 · newsletter

Enterprise Grade ML

by Shioulin · At Cloudera Fast Forward, one of the mechanisms we use to tightly couple machine learning research with application is through application development projects for both internal and external clients. The problems we tackle in these projects are wide ranging and cut across various industries; the end goal is a production system that translates data into business impact. What is Enterprise Grade Machine Learning? Enterprise grade ML, a term mentioned in a paper put forth by Microsoft, refers to ML applications where there is a high level of scrutiny for data handling, model fairness, user privacy, and debuggability. more
Apr 1, 2020 · post

Bias in Knowledge Graphs - Part 1

by Keita · Introduction This is the first part of a series to review Bias in Knowledge Graphs (KG). We aim to describe methods of identifying bias, measuring its impact, and mitigating that impact. For this part, we’ll give a broad overview of this topic. image credit: Mediamodifier from Pixabay Motivation Knowledge graphs, graphs with built-in ontologies, create unique opportunities for data analytics, machine learning, and data mining. They do this by enhancing data with the power of connections and human knowledge. more

Popular posts

Oct 30, 2019 · newsletter
Exciting Applications of Graph Neural Networks
Nov 14, 2018 · post
Federated learning: distributed machine learning with data locality and privacy
Apr 10, 2018 · post
PyTorch for Recommenders 101
Oct 4, 2017 · post
First Look: Using Three.js for 2D Data Visualization
Aug 22, 2016 · whitepaper
Under the Hood of the Variational Autoencoder (in Prose and Code)
Feb 24, 2016 · post
"Hello world" in Keras (or, Scikit-learn versus Keras)


In-depth guides to specific machine learning capabilities


Machine learning prototypes and interactive notebooks

Explain BERT for Question Answering Models

Tensorflow 2.0 notebook to explain and visualize a HuggingFace BERT for Question Answering model.

NLP for Question Answering

Ongoing posts and code documenting the process of building a question answering model.

Interpretability Revisited: SHAP and LIME

Explore how to use LIME and SHAP for interpretability.


Refractor predicts churn probabilities for telecom customers and shows which customer attributes contribute to those predictions.


Cloudera Fast Forward is an applied machine learning reseach group.
Cloudera   Blog   Twitter