INDEX

Explanations

phrases describing unauthorized access or breaches

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Configuration

Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post

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

Negative Logits

ibraries

-0.16

iba

-0.16

ÑĥÑī

-0.15

ooks

-0.15

ÏĦÏģÎ±

-0.14

ebo

-0.14

gamber

-0.14

abet

-0.14

âĶĺ

-0.14

Locator

-0.14

POSITIVE LOGITS

 label

0.21

big

0.20

 crosses

0.19

 note

0.18

 neon

0.18

 crossed

0.17

 sign

0.17

 Label

0.17

 Crossing

0.17

 signs

0.16

Activations Density 0.024%

phrases describing unauthorized access or breaches

No Known Activations