Reproductions
The Reproduction Gallery collects end-to-end reproductions of published mechanistic interpretability results, implemented with Murano pipelines. Each entry in the gallery documents the paper, the experimental setup, and the results obtained, and comes with a runnable notebook.
Browse the galleryWhy reproductions?
Section titled “Why reproductions?”Interpretability moves fast, and papers rarely agree on APIs, data formats, or evaluation protocols. The gallery exists so that anyone can:
- Run published experiments in a few lines — Murano’s unified
PipelineAPI collapses activation recording, probing, steering, and evaluation into declarative composable steps. - Compare results across papers — same dataset format, same model loader, same metrics.
- Extend and remix — every reproduction ships as a notebook you can fork and modify.
What’s in a reproduction?
Section titled “What’s in a reproduction?”Each gallery entry contains:
| Section | What it covers |
|---|---|
| Paper metadata | Venue, authors, BibTeX, links to arXiv / OpenReview / original code |
| TL;DR | A 2–3 sentence summary of the core claim |
| Abstract | Verbatim paper abstract for context |
| Reproducing with Murano | Step-by-step walkthrough with runnable code |
| Key results | A table comparing paper vs. reproduction numbers |
| Citation | BibTeX entry for the original work |
Have something to add?
Section titled “Have something to add?”If you’ve reproduced an interpretability paper with Murano — or you’re using Murano in your own research and want to share a reference implementation — we’d love to feature it.
Submit a reproduction →