INDEX

Explanations

references to position or placement

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

 above

-0.24

 Above

-0.23

 ABOVE

-0.22

Above

-0.20

bove

-0.16

lite

-0.15

à¸©

-0.15

ialis

-0.15

ä»¥ä¸Ĭ

-0.15

 oben

-0.15

POSITIVE LOGITS

 âĨĵ

0.26

neath

0.24

âĨĵ

0.22

/left

0.21

0.20

-left

0.20

/right

0.20

-average

0.19

éĿ¢çļĦ

0.19

 ÄĳÃ¢y

0.18

Activations Density 0.028%

references to position or placement

No Comments

No Known Activations