Open Artifacts​

Tools, datasets, and checklists we release.

We publish the datasets, evaluation methods, and teaching materials behind our papers so others can reuse, audit, and build on them. Everything here is openly licensed on Zenodo and GitHub.

Method · LLM evaluation · 2026

ArC: Argument-Based Consistency
Software and datasets that measure whether an LLM’s free-form toxicity explanations hold together as a logical argument.
 

Zenodo · GitHub

Method · Worksheet · 2026

RaR: Reasoning about Reasoning
A worksheet for the reflective, accountable use of LLM reasoning in HCI research: when, why, and how researchers rely on model-generated reasoning.
 

Zenodo · GitHub

Method · Worksheet · 2025

AsAP: Assumption-as-Premise
A framework that turns implicit machine-learning assumptions into documented, inspectable, and contestable premises.
 

Zenodo · GitHub

Each artifact accompanies a peer-reviewed paper. See Publications for the full record. More of the lab’s releases are on Zenodo.