INDEX

Explanations

terms related to reactions and responses

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

ÐµÐ·

-0.15

cott

-0.15

sert

-0.14

ser

-0.14

 ÐĹÐ°Ñħ

-0.14

plat

-0.14

 rencontres

-0.14

MENT

-0.14

Ã¶z

-0.14

RIPT

-0.14

POSITIVE LOGITS

ivate

0.35

aries

0.31

ively

0.24

ary

0.20

iveness

0.18

/respond

0.18

/response

0.17

uate

0.17

ives

0.17

ero

0.16

Activations Density 0.013%

terms related to reactions and responses

No Comments

No Known Activations