INDEX

Explanations

phrases encouraging the reader to explore or discover additional content or resources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 itself

-0.15

vented

-0.15

 Himself

-0.14

ãģĭãĤĬ

-0.14

ä¹ĭ

-0.14

osy

-0.14

áº©u

-0.14

edback

-0.14

Å¥

-0.14

Ã¨ge

-0.14

POSITIVE LOGITS

how

0.27

 some

0.23

 what

0.20

some

0.19

 below

0.19

why

0.18

 other

0.18

www

0.17

 latest

0.17

 cÃ³mo

0.17

Activations Density 0.046%

phrases encouraging the reader to explore or discover additional content or resources

No Comments

No Known Activations