INDEX

Explanations

why followed by a verb

The neuron activates on the words in rhetorical “why … when …” questions, e.g. “why choose old version when ….”

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

 měli

-1.16

 firent

-1.07

瓛

-1.05

ПРИ

-1.04

 unſ

-1.04

!*\

-1.04

面白かった

-1.03

 fono

-1.02

妁

-1.02

umenter

-1.01

POSITIVE LOGITS

 then

1.78

1.48

now

1.16

1.14

1.10

let

1.09

 anymore

1.02

 anyone

1.02

 Then

1.00

Activations Density 0.015%