INDEX

Explanations

phrases indicating opinions or evaluations related to media or performances

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

UBLE

-0.16

aders

-0.15

 Bene

-0.14

 certain

-0.14

 Lehr

-0.14

-0.13

eses

-0.13

 Honest

-0.13

Nar

-0.13

urat

-0.13

POSITIVE LOGITS

 shallow

0.22

 nons

0.21

 irrelevant

0.20

 meaningless

0.20

 artificial

0.20

 fabricated

0.20

 pointless

0.19

 useless

0.19

æ¯«

0.19

 arbitrary

0.18

Activations Density 0.040%

phrases indicating opinions or evaluations related to media or performances

No Comments

No Known Activations