LLM+RAG based Question Answering

Retrieval Augmented Generation (RAG) seems to be quite popular these days. Along the wave of Large Language Models (LLM’s), it is one of the popular techniques to get LLM’s to perform better on specific tasks such as question answering on in-house documents. Some time ago, I played on a Kaggle competition that allowed me to […]

LLM at Home

Large Language Models (LLMs) are the greatest hype recently. It started with ChatGPT but now there are numerous open source, or openly available (open access?), models to try locally. Mostly on Huggingface. Building (training) an entire LLM model requires quite massive resources, but the opportunity of using these available (pre-trained) models is very interesting and […]

Algorithm Test Engineering Part 2: Machine Learning

If we don’t exactly know what we are testing, how can we test? In my previous algorithm test engineering article, I discussed the testing and analysis of more classical algorithms, such as binary search. While the overall testing of algorithms can be complicated, most classical algorithms can be described as having quite clearly defined inputs, outputs, […]

Algorithm Test Engineering: Exploratory Job Analysis

Recently I had a discussion about what it means to test an algorithm, or what it means to do algorithm test engineering. I couldn’t quite come up with a convincing definition for myself. At a basic level, maybe you figure out some basic rules of the algorithm, give it some input and output. Check the results. But I believe there is more to it, and this is what I explore in this article.

Mapping Ring Signatures and Stealth Addresses in Monero

Monero employs many techniques to enhance the cryptocurrency users privacy. In this article, I take a look at two of these, Ring Signatures for sender anonymity, and Stealth Addresses for receiver anonymity.

Merkle Trees: Concepts and Use Cases

ms including Bitcoin, Amazon’s Dynamo DB, and the ZFS filesystem. The basic concept is quite simple, but some of the clever applications are not so obvious.

A Look at Precision, Recall, and F1-Score

Precision, Recall, and F1-Score. These relate to getting a finer-grained idea of how well a classifier is doing, as opposed to just looking at overall accuracy. In this article, I provide a bit deeper look at each.

Metamorphic Testing of Machine-Learning Based Systems

aditional software. With traditional software, the specification and its relation to the implementation is typically quite explicit. With more complex machine learning-based system, this relation is harder to explicitly define. This makes testing them more complicated. In this article, I present an updated version of my earlier work on using metamorphic testing for ML based systems.

Zero Knowledge Proofs: Example with Pedersen Commitments in Monero

An Explanation Between Cartoons and Greek Symbols If you spend a bit more time reading about cryptocurrencies, blockchains, or many other related technologies, you likely run into the term zero knowledge proofs (ZKP). To me, the term sounds like a paradox. How can you prove anything with zero knowledge? How do you event know what to […]

Building an NLP classifier: Example with Firefox Issue Reports

results these days. Libraries such as Keras, PyTorch, and HuggingFace NLP make the application of the latest research and models in the area a (relatively) easy task. In this article, I implement and compare two different NLP based classifier model architectures using the Firefox browser issue tracker data.