INDEX

Explanations

terms associated with spreading information or awareness

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ettle

-0.17

İ

-0.15

ima

-0.15

urs

-0.15

oon

-0.15

ff

-0.14

Preferences

-0.14

cul

-0.13

imitives

-0.13

pii

-0.13

POSITIVE LOGITS

sheet

0.30

heets

0.27

 throughout

0.23

 widely

0.21

shirt

0.21

 wide

0.21

sheets

0.20

wid

0.20

Wide

0.20

vir

0.19

Activations Density 0.038%

terms associated with spreading information or awareness

No Comments

No Known Activations