INDEX

Explanations

actions done

This neuron detects emphatic confirmation phrases like “that’s exactly what…” or “exactly what…” in the text.

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

of

-1.60

is

-1.41

 ofere

-1.28

urem

-1.26

 других

-1.21

 अन्य

-1.20

太多的

-1.18

 другие

-1.16

 introduce

-1.15

 دیگر

-1.15

POSITIVE LOGITS

but

1.57

one

1.34

 when

1.32

 some

1.25

 with

1.25

 three

1.24

 what

1.19

when

1.19

 this

1.17

广西

1.16

Activations Density 0.034%