INDEX

Explanations

function/action followed by preposition

The neuron strongly responds to multisyllabic content words (longer lexical nouns and verbs) rather than to short function words.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ícil

-1.09

 loosely

-1.05

 برای

-1.05

}$.

-0.98

ParallelGroup

-0.98

對於

-0.96

 mesmo

-0.94

 those

-0.91

 broadly

-0.91

 для

-0.91

POSITIVE LOGITS

in

1.71

on

1.15

法则

1.10

 inicialmente

1.02

 with

0.98

 using

0.97

 طويلة

0.94

at

0.93

 according

0.93

 именно

0.91

Activations Density 0.230%