English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Understanding the function of individual neurons within language models isessential for mechanistic interpretability research. We propose $\textbf{Neuronto Graph (N2G)}$, a tool which takes a neuron and its dataset examples, andautomatically distills the neuron's behaviour on those examples to aninterpretable graph. This presents a less labour intensive approach tointerpreting neurons than current manual methods, that will better scale thesemethods to Large Language Models (LLMs). We use truncation and saliency methodsto only present the important tokens, and augment the dataset examples withmore diverse samples to better capture the extent of neuron behaviour. Thesegraphs can be visualised to aid manual interpretation by researchers, but canalso output token activations on text to compare to the neuron's ground truthactivations for automatic validation. N2G represents a step towards scalableinterpretability methods by allowing us to convert neurons in an LLM tointerpretable representations of measurable quality.

N2G: A Scalable Approach for Quantifying Interpretable Neuron Representations in Large Language Models