INDEX

Explanations

the preposition "for."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

for

-0.21

 dafÃ¼r

-0.19

 untuk

-0.18

 fÃ¼r

-0.17

 Î³Î¹Î±

-0.17

 iÃ§in

-0.16

 voor

-0.16

 Ø¨Ø±Ø§ÛĮ

-0.15

for

-0.15

erta

-0.15

POSITIVE LOGITS

 sake

0.40

 purposes

0.34

bidden

0.31

geries

0.31

-profit

0.30

aging

0.30

 instance

0.28

/by

0.28

king

0.27

ays

0.27

Activations Density 0.721%

the preposition "for."

No Comments

No Known Activations