INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ucz

-0.20

lew

-0.15

 Cone

-0.15

oday

-0.15

eled

-0.14

æµ¦

-0.14

.yang

-0.13

 –↵↵

-0.13

.none

-0.13

cale

-0.13

POSITIVE LOGITS

Kh

0.16

uppy

0.15

dialogs

0.15

Kh

0.15

gang

0.15

mime

0.14

kh

0.14

 Demp

0.14

hek

0.14

oze

0.14

Activations Density 0.000%

No Known Activations

This feature has no known activations.