INDEX

Explanations

discuss or give an action

The neuron fires on first‐person‐plural future‐tense constructions (e.g. “we will…,” “we’re going to…”) that signal the author’s upcoming action.

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

 this

-1.85

 their

-1.83

in

-1.75

on

-1.70

by

-1.48

now

-1.38

 where

-1.26

to

-1.21

nemia

-1.18

 then

-1.17

POSITIVE LOGITS

 autres

1.41

1.38

 externe

1.34

 autre

1.34

 quelconque

1.34

utsal

1.33

↵↵

1.33

 OGSÅ

1.30

lain

1.30

 repuesto

1.28

Activations Density 0.089%