Year: 2023
Type: preprint
Abstract: Recent work has shown that computation in language models may be human-understandable, with successful efforts to localize and intervene on both single-unit features and input-output circuits. Here, we introduce an approach which extends causal mediation experiments to automatically identify model components responsible for performing a specific subtask by solely specifying a set of \textit{desiderata}, or causal attributes of the model components executing that subtask. As a proof of concept, we apply our method to automatically discover shared \textit{variable binding circuitry} in LLaMA-13B, which retrieves variable values for multiple arithmetic tasks. Our method successfully localizes variable binding to only 9 attention heads (of the 1.6k) and one MLP in the final token's residual stream.
Source: arXiv (Cornell University)
Authors: Davies, Xander, Nadeau, Max, Prakash, Nikhil, Shaham, Tamar Rott, Bau, David
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funder: Open Philanthropy Project
OpenAlex ID: W4383996906