INDEX

Explanations

references to social media platforms, particularly Twitter and Facebook

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

 Facebook

-0.17

elow

-0.16

 Websites

-0.15

inalg

-0.15

 Alphabet

-0.14

 facebook

-0.14

epend

-0.14

perimental

-0.14

onse

-0.14

FACE

-0.14

POSITIVE LOGITS

 account

0.40

 page

0.30

 feed

0.29

 accounts

0.29

account

0.28

 Account

0.25

 presence

0.25

 handle

0.25

-account

0.24

Account

0.23

Activations Density 0.037%

references to social media platforms, particularly Twitter and Facebook

No Known Activations