# Modular addition without black-boxes: Compressing explanations of MLPs that compute numerical integration

arXiv: https://arxiv.org/abs/2412.03773 · HTML: https://mlsanity.com/paper/2412.03773/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability.
- Builds on 6 of its 10 references that are in the map.
- First posted on arXiv 4 Dec 2024.

## Placed in

- Mechanistic interpretability / [Circuits and mechanistic analysis](https://mlsanity.com/mechinterp/circuits-and-mechanistic-analysis/)

## Builds on (most cited first, 6 of 6)

- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [Toy Models of Superposition](https://mlsanity.com/paper/2209.10652/) (2022)
- [Progress measures for grokking via mechanistic interpretability](https://mlsanity.com/paper/2301.05217/) (2023)
- [Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models](https://mlsanity.com/paper/2403.19647/) (2024)
- [The Clock and the Pizza: Two Stories in Mechanistic Explanation of Neural Networks](https://mlsanity.com/paper/2306.17844/) (2023)
- [Compact Proofs of Model Performance via Mechanistic Interpretability](https://mlsanity.com/paper/2406.11779/) (2024)

## Built on by (most cited first, 3 of 3)

- [Path Channels and Plan Extension Kernels: a Mechanistic Description of Planning in a Sokoban RNN](https://mlsanity.com/paper/2506.10138/) (2025)
- [On the geometry and topology of representations: the manifolds of modular addition](https://mlsanity.com/paper/2512.25060/) (2025)
- [Interactions Between Crosscoder Features: A Compact Proofs Perspective](https://mlsanity.com/paper/2606.09940/) (2026)

## Abstract (arXiv)

The goal of mechanistic interpretability is discovering simpler, low-rank algorithms implemented by models. While we can compress activations into features, compressing nonlinear feature-maps -- like MLP layers -- is an open problem. In this work, we present the first case study in rigorously compressing nonlinear feature-maps, which are the leading asymptotic bottleneck to compressing small transformer models. We work in the classic setting of the modular addition models, and target a non-vacuous bound on the behaviour of the ReLU MLP in time linear in the parameter-count of the circuit. To study the ReLU MLP analytically, we use the infinite-width lens, which turns post-activation matrix multiplications into approximate integrals. We discover a novel interpretation of} the MLP layer in one-layer transformers implementing the ``pizza'' algorithm: the MLP can be understood as evaluating a quadrature scheme, where each neuron computes the area of a rectangle under the curve of a trigonometric integral identity. Our code is available at https://tinyurl.com/mod-add-integration.