INDEX

Explanations

file paths and code structures

The neuron is sensitive to numeric tokens or digit sequences (e.g. version numbers, IDs, counts).

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

 介绍

-1.10

aile

-1.09

 Grüßen

-1.08

と思っています

-1.07

してしまう

-1.05

is

-1.02

Reviewers

-1.02

excellents

-1.01

 haupts

-1.00

POSITIVE LOGITS

alism

1.26

romeda

1.24

₈

1.18

henswürdigkeiten

1.17

),

1.16

♭

1.16

ogues

1.15

chestra

1.15

$\

1.14

-\

1.13

Activations Density 0.039%