INDEX

Explanations

instances of numerical data and statistical scores

New Auto-Interp

Configuration

Features

65,536

Data Type

float32

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

gated

Context Size

1,024

Dataset

Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Manson

-0.17

åĪ·

-0.16

ĩ¼

-0.16

eri

-0.15

 Mellon

-0.15

 Eleanor

-0.15

onet

-0.15

 Caldwell

-0.14

Ø±ÙĪØ´

-0.14

 Barcl

-0.14

POSITIVE LOGITS

0.36

0.34

0.32

0.27

0.25

0.24

Activations Density 0.023%