INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

rana

-0.16

asant

-0.15

aterno

-0.15

olib

-0.14

Ctrls

-0.14

atorial

-0.14

StackNavigator

-0.14

amarin

-0.14

ousel

-0.14

prep

-0.14

POSITIVE LOGITS

 Boyle

0.17

pins

0.16

odel

0.16

Ð¾Ð»Ð¾Ñģ

0.16

rend

0.15

ãĥ³ãĥĩãĤ£

0.15

.hu

0.15

 profit

0.15

rends

0.15

 Downs

0.15

Activations Density 0.000%

No Known Activations

This feature has no known activations.