# Mathematical Models of Computation in Superposition

Preprint · Kaarel Hänni, Jake Mendel, Dmitry Vaintrob, Lawrence Chan

- Archive id: arxiv:2408.05451
- Published: 2024-08-10
- arXiv: https://arxiv.org/abs/2408.05451
- Citations: 1
- Source URL: https://arxiv.org/abs/2408.05451
- PDF: https://arxiv.org/pdf/2408.05451
- Provenance: arxiv

## Abstract

Superposition -- when a neural network represents more \`\`features'' than it has dimensions -- seems to pose a serious challenge to mechanistically interpreting current AI systems. Existing theory work studies \\emph{representational} superposition, where superposition is only used when passing information through bottlenecks. In this work, we present mathematical models of \\emph{computation} in superposition, where superposition is actively helpful for efficiently accomplishing the task. We first construct a task of efficiently emulating a circuit that takes the AND of the $\\binom{m}{2}$ pairs of each of $m$ features. We construct a 1-layer MLP that uses superposition to perform this task up to $\\varepsilon$-error, where the network only requires $\\tilde{O}(m^{\\frac{2}{3}})$ neurons, even when the input features are \\emph{themselves in superposition}. We generalize this construction to arbitrary sparse boolean circuits of low depth, and then construct \`\`error correction'' layers that allow deep fully-connected networks of width $d$ to emulate circuits of width $\\tilde{O}(d^{1.5})$ and \\emph{any} polynomial depth. We conclude by providing some potential applications of our work for interpreting neural networks that implement computation in superposition.

_Abstract reproduced under CC0-1.0 (arXiv descriptive metadata)._

Topics: [Mechanistic Interpretability](https://space-kitty.com/topics/interpretability)

---
Space Kitty Archive · https://space-kitty.com · API: https://space-kitty.com/api/v1 · MCP: https://space-kitty.com/mcp · Premium: $0.20/request via x402 (Base Sepolia (testnet)). Metadata: arXiv (CC0), OpenAlex (CC0), Europe PMC (CC BY/CC0 works), ClinicalTrials.gov.