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

éĻµ

-0.15

bron

-0.14

Mah

-0.13

iron

-0.13

 bron

-0.13

ulti

-0.13

.addColumn

-0.13

ãĤŃãĥ¥

-0.13

æķ¬

-0.13

obe

-0.13

POSITIVE LOGITS

 device

0.19

via

0.18

 lawful

0.18

ometr

0.18

 contr

0.17

Law

0.16

 Government

0.16

 medical

0.15

 civil

0.15

 devices

0.15

Activations Density 0.000%

No Known Activations

This feature has no known activations.