Explore
Explore
Interested in learning more about the things we explore, code, and build? Maybe you’re curious about how the confluence of Open Source technology and applied research impacts national security and democratically aligned values? Or perhaps you’re interested in collaborating with us on any of our active research projects? Find out more about what we are doing and learning below.
Read
Have a broad look at the tech we play with, insights we learn along the way, and even some behind-the-scenes findings from our projects.
Is your AI a “Clever Hans”?
"Clever Hans," the horse mathematician, was a worldwide sensation in 1904, but his uncanny abilities were eventually debunked. As it turns out, Clever Hans had accidentally been trained to recognize his owners’…
IQT Labs releases audit report of RoBERTa, an open source large language model
Large Language Models (LLMs) are tremendously powerful, but also concerning, in part because of their potential to generate offensive, stereotyped, and racist text. Since LLMs are trained on extremely large text datasets…
Interrogating RoBERTa: A Case Study in Cybersecurity
Found It! How we discovered a security hole in a popular software actively used by millions of people around the world? This blog details an unexpected discovery in our audit of a…
Eating Every Last Slice of Your Raspberry Pi – GPU Offload on a Pi4
IQT Labs has been experimenting using Raspberry Pi4s and Software Defined Radios (SDR) as an on-the-go platform, through the GamutRF project. Learn how we made GamutRF, running on a Pi4, more responsive…
Saisiyat is where it is at! Auditing multi-lingual AI models
How do you audit an AI model and what can you expect to learn from it? This blog explores auditing multi-lingual AI models.
Interrogating RoBERTa: Inside the challenge of learning to audit AI models and tools
This post offers a look inside the challenge of learning to audit AI models and tools, exploring an audit of RoBERTa, an open source, pretrained Large Language Model.
The Promises and Perils of Adversarial Camouflage
An introduction to the adversarial camouflage project, focusing on how effective legacy patches truly are.
Deep Geolocation with Satellite Imagery of Ukraine
As the Russian invasion of Ukraine plays out before the world on social media, it’s critical to verify the places where the photos/videos were recorded. This blog explores our attempt to geolocate…
Fusing Without Confusing
Humans use all available senses to identify the world around them, a concept often called "sensor fusion" in Machine Learning. This blog focuses on how video and audio sensors can help compensate…
Fusing Without Confusing - A Unique Data Processing Pipeline for VoxCeleb2
We developed a DataLoader that can feed audio and video into a single machine learning model without additional processing - the first of its kind! Fusing Without Confusing uses multi-modal deep learning…
Explore AI Sonobuoy: IQT Labs Throws TinyML in the Deep End
What happened when we dropped one of our latest creations out of a boat? Read on to learn about the story of the AI Sonobuoy — a low-cost hydrophone that uses AI…
Where in the World, Part 3: Further Analysis of Cross-view Image
Why do deep learning models for cross-view image geolocalization (CVIG) behave as they do? In this final blog from our three-part series exploring CVIG, we explore several aspects of that question.
Where in the World, Part 2: A Deep Learning Model for Cross-view Image
Learn how we implemented a deep learning model and used it with our cross-view image geolocalization (CVIG) dataset to see what boosts performance and what doesn't.
Introducing Atlas: A Tool for Visualizing Deep Tech Relationships at Scale
Map out your startup ecosystem. Analyze your citation and investing networks. A graphic of your connected nodes via Atlas can map your anything.
Where in the World, Part 1: A New Dataset for Cross-view Image
Suppose that a photograph has surfaced on social media, or through some dubious origin, and you want to know: where was it really taken? This post explores our new CVIG dataset using…
AI Assurance: Do deepfakes discriminate?
Do deepfakes discriminate? We explore this concept in the fifth blog post of our AI Assurance series, discussing the bias portion of our audit of the deepfake detection tool, Fakefinder.
Open Source Software Nutrition Labels: Scaling Up with Anaconda Python
Our first blog in this series introduced our Nutrition Label concept, focused on providing opensource software transparency. This final post in the series explores how we scaled up this Open Source Software…
AI Assurance: An “ethical matrix” for FakeFinder
With so many ways that AI can fail, how do we know when an AI tool is “ready” and “safe” for deployment? Our fourth blog post in our AI Assurance series explores…
Measuring Global Bio-Power: The Strengths and Limits of Using Open Source Software Metadata to Analyze Worldwide Technical Trends
Interested in global technology trends but frustrated by the shortcomings of patent and citation data? Learn how we analyzed global bioinformatics trends using opensource software metadata, and see what tactics you can…
Open Source Software Nutrition Labels: An AI Assurance Application
Since software supply chains are often opaque and complex, we recently explored the concept of an Open Source Software “Nutrition Label” as part of our AI Assurance work.
AI Assurance: A deep dive into the cybersecurity portion of our FakeFinder audit
What happened when we dove into the cybersecurity portion of our FakeFinder audit? Read on to find out.
Synthesizing Robustness YOLTv4 Results Part 2: Dataset Size Requirements and Geographic Insights
In previous blogs we discussed the dataset and initial aggregate results for our “Synthesizing Robustness” project. Read further to learn about detailed results for object detection models, focusing on geographic disparities and…
Watch & Listen
Our videos and podcasts highlight projects and topics of interest.
IQT Explains: The Importance of Bias Testing in AI Assurance
An IQT Podcast episode exploring how we test and assess AI technology to minimize unwanted biases and consider legalities and ethics. Listen in if you want to discover how AI technology is…
IQT Explains: AI Assurance
What happened when we audited a deepfake detection tool called FakeFinder? Our AI Assurance series continues with a podcast detailing findings from our audit on this Open Source tool that predicts whether…
IQT Labs Presents FakeFinder, an Open Source Tool to Help Detect Deepfakes
How do you know if a video clip is representative of reality, or the person you’re talking to on the other side of a video call is who they appear to be?…
Building a Custom Object Detection Model
Interested in learning how to use machine learning to detect different objects? This tutorial walks through the process of using the Tensorflow 2.x Object Detection API to build an airplane detector.
Viziflu - Visualizing Uncertainty in Crowd-sourced Influenza Forecasts
In 2018, IQT Labs began working with the Centers for Disease Control & Prevention to create Viziflu, a tool to support infectious disease forecasting. Learn more about this web-based software tool that…
Build
Build your own version of something we created at IQT Labs! Check out our GitHub Repos here and let us know how it goes.
ViziFlu GitHub Repo
An open-source visualization tool to help make multi-model seasonal influenza forecasts more actionable for decision-makers.
Daedulus GitHub Repo
This project explores mechanisms to make informed decisions from a variety of response options that can be used to mitigate attack effectiveness by steering malicious actors to domains that the defender can…
BirdsEye GitHub Repo
The BirdsEye RL/RF project enables localization of mobile radio frequency targets, e.g., drone operators, via commercial off-the-shelf sensors.
FakeFinder GitHub Repo
FakeFinder builds a modular framework for evaluating various deepfake detection models, offering a web application as well as API access for integration into existing workflows.
Atlas GitHub Repo
Atlas is a visualization tool for creating customizable, interactive graphs in a browser using Python and Dash Cytoscape.
Investigate
Get deep into the technical details of our explorations, projects, and findings.
AI Assurance Audit of RoBERTa, an Open source, Pretrained Large Language Model
Large Language Models (LLMs) have been in the news lot recently due to their tremendous power, but also concerns around their potential to generate offensive text. It's difficult to know how LLMs…
AI Assurance Audit of FakeFinder
What we found when we audited a deepfake detection tool called FakeFinder.
Synthesizing Robustness Report
Is your synthetic data letting you down? This report explores whether generative models can increase the utility of synthetic data for rare object detection.
Copyright © 2022 · IQT Labs LLC. All Rights Reserved.
