INDEX

Explanations

terms that indicate significance or importance

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

istrovstvÃŃ

-0.16

andaÅŁ

-0.16

 naken

-0.16

SCII

-0.15

Äįin

-0.14

 Ø§ÙĦØ±Ø¦ÙĬØ³

-0.14

lope

-0.14

.mc

-0.14

.lv

-0.14

optgroup

-0.14

POSITIVE LOGITS

 role

0.66

 roles

0.52

role

0.51

 Role

0.51

Role

0.44

-role

0.43

 Roles

0.41

_role

0.41

 ÑĢÐ¾Ð»ÑĮ

0.40

roles

0.40

Activations Density 0.037%

terms that indicate significance or importance

No Comments

No Known Activations