INDEX

Explanations

conversely, converse

The neuron fires on occurrences of “converse” (including “Conversely”) as a discourse marker indicating contrast.

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

We

-2.31

啧

-2.31

你来

-2.30

 expresión

-2.28

 amable

-2.23

 видимо

-2.23

 цветами

-2.17

 Insets

-2.16

atief

-2.14

 Vereinigten

-2.13

POSITIVE LOGITS

2.75

2.72

这种

2.58

2.50



2.48

這

2.45

她

2.42

一

2.36

每次

2.28

这

2.27

Activations Density 0.002%