INDEX

Explanations

don't have to + verb

This neuron detects negative necessity constructions—phrases like “don’t have to …,” highlighting words in “do not have to” + verb contexts.

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

μως

-1.66

if

-1.48

chure

-1.40

無事に

-1.39

这一幕

-1.39

pearance

-1.36

ARROLLO

-1.33

 pokud

-1.33

,</

-1.31

 palab

-1.30

POSITIVE LOGITS

).

1.30

直至

1.24

১

1.23

一次

1.22

bzw

1.20

 about

1.20

Ajax

1.20

 один

1.20

 beziehungs

1.19

1.18

Activations Density 0.022%