INDEX

Explanations

phrases involving repetition or cycles

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

addtogroup

-0.17

ledo

-0.15

lesen

-0.15

edBy

-0.14

ÑįÑĤ

-0.14

tal

-0.14

ibt

-0.14

ë¡ľëĤĺ

-0.14

aka

-0.13

Ã¥l

-0.13

POSITIVE LOGITS

-and

0.39

and

0.28

and

0.25

_and

0.24

and

0.22

 Ð¸

0.21

.and

0.21

åĴĮ

0.20

 vÃł

0.20

 åĴĮ

0.19

Activations Density 0.085%

phrases involving repetition or cycles

No Comments

No Known Activations