Skip to content

2026

Relevance-bench

This is a follow up to my post on using Jev for pseudo-relevance judgments on TREC DL 2023.

This experiment felt more useful as a little benchmark that I can extend.

Check out the code for this post here. For the data used, see my dataset on Hugging Face here.

Pseudo-relevances with Jev

The world is abuzz with excitement at the new "system one" models being pumped out right now. In this post, I take a look at how good these models are at generating relevancy scores on the TREC {0, 1, 2, 3} scale.

The code used to create the results in this blog can be found here.

Exploring Kendall's \(\tau\)

For one reason or another, I've been working with Kendall's \(\tau\) lately. This post explores a little bit of my excitement with the statistic and some possible extensions. Disclaimer: I am not a statistician.

The code used to created the plots in this blog can be found here.

Roguelites and gambling

I've been show and tell heavy lately. So, here's a lighter blog post about my gaming habits and an examination of the roguelite genre. It comes very close to saying something... almost.

Apologies in advance for my self-reflection on my own history of gaming and current habits.

Zierra part 1

Back with some non-Python content. This is the first installment of my journey to implement Tierra in Zig (Zierra) and to use LLMs for learning.

Rogue-Bench

After a couple months of work, she's finally done. Rogue-Bench is my contribution to the already overflowing list of LLM benchmarks.

This is an extension of my previous work here. So check that out first for some context.

Death by dog food

I've been slacking on blog posts. So this one is a filler episode at best.

I learned the term dogfooding recently and it really resonated.

I'm overloading the term a little bit in this blog but, I hope the story still works.

Learning Zig part 2

Here is my second entry in my "learning Zig" series. Likely the final.

I feel I have seen enough syntax and am ready to build something.

For my first post on this topic, see here.