Murano
Docs Reproduction Gallery

Reproduction Gallery

End-to-end reproductions of published interpretability results, implemented with Murano pipelines. Each entry includes a runnable notebook, paper reference, and walkthrough.

ICLR 2023 circuitsattentionpath patching

Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 Small

Wang, Variengien, Conmy, Shlegeris, Steinhardt

Reverse-engineers how GPT-2 small identifies the indirect object, localizing the name-mover, negative name-mover, and S-inhibition heads. Reproduced with Murano path patching, attention analysis, and OV circuits.

View reproduction →
Paper ↗ Code ↗
arXiv 2023 probingrepresentationssteering

The Geometry of Truth: Emergent Linear Structure in LLM Representations of True/False Datasets

Marks, Tegmark

Shows that the truth of a factual statement is encoded as a single linear direction in the residual stream: true and false statements separate, a mass-mean probe reads the direction, it generalizes across topics, and adding it flips the model's judgement. Reproduced with Murano recording, probing, and interventions.

View reproduction →
Paper ↗ Code ↗
ICLR 2024 in-context learningcausal mediationsteering

Function Vectors in Large Language Models

Todd, Li, Sen Sharma, Mueller, Wallace, Bau

Shows that a small set of attention heads carry a compact, portable representation of an in-context task: summing their outputs gives a function vector, and adding it to a zero-shot prompt makes GPT-J perform the task. Reproduced with Murano recording, causal mediation, and interventions.

View reproduction →
Paper ↗ Code ↗
TMLR 2026 probingrepresentationsgeometry

Shape Happens: Automatic Feature Manifold Discovery in LLMs via Supervised Multi-Dimensional Scaling

Tiblias, Bigoulaeva, Niu, Balloccu, Gurevych

Introduces SMDS, a framework for discovering and comparing geometric hypotheses about how LLMs organize features in latent space. Shows that feature manifolds form circles, chains, and clusters reflecting the semantic structure of the encoded concept.

View reproduction →
Paper ↗ Code ↗
Your reproduction here. Submit a notebook and we'll feature it in the gallery.
Murano · UKP Lab, TU Darmstadt GitHub