INDEX

Explanations

references to audience perception and engagement with characters and narratives

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

Not in Any Lists

No Comments

Negative Logits

quat

-0.17

defgroup

-0.17

eday

-0.15

ersh

-0.15

ÙĤØ§ÙĦ

-0.15

itel

-0.15

 joking

-0.15

borg

-0.14

inde

-0.14

ulas

-0.14

POSITIVE LOGITS

 invest

0.26

 investment

0.24

 invested

0.23

 suspend

0.23

 viewer

0.22

 suspension

0.22

suspend

0.22

 rooting

0.21

 investing

0.21

vic

0.21

Activations Density 0.103%

references to audience perception and engagement with characters and narratives

No Comments

No Known Activations