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

æµ©

-0.15

udit

-0.15

ï¸ı

-0.15

ÐµÐ´Ð°Ð³

-0.14

çĩŁ

-0.14

ursal

-0.14

.AddSingleton

-0.14

yntax

-0.14

nock

-0.13

.appspot

-0.13

POSITIVE LOGITS

 slavery

0.27

 slave

0.25

 slaves

0.25

 Slave

0.23

_slave

0.21

 Northern

0.21

Slave

0.21

slave

0.20

 Southern

0.20

å¥´

0.18

Activations Density 0.000%

No Known Activations

This feature has no known activations.