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

à¸²à¸ĩ

-0.15

Bis

-0.15

à¸Ļà¸²à¸Ķ

-0.14

Ð»ÐµÑĩ

-0.14

uger

-0.14

ãĤ·ãĤ§

-0.14

presso

-0.13

orna

-0.13

.smart

-0.13

"';

-0.13

POSITIVE LOGITS

 Disp

0.18

 Paul

0.17

 sola

0.17

 Christ

0.17

Yah

0.16

 DISP

0.16

 Brothers

0.15

 contention

0.15

beit

0.15

 Brother

0.15

Activations Density 0.000%

No Known Activations

This feature has no known activations.