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

 láº¡c

-0.15

anni

-0.15

TRL

-0.14

ç¿Ķ

-0.14

å±Ĭ

-0.14

erea

-0.14

kad

-0.14

autop

-0.13

achelor

-0.13

aska

-0.13

POSITIVE LOGITS

AJ

0.63

AJ

0.56

Aj

0.52

aj

0.51

Aj

0.45

aj

0.44

PJ

0.41

CJ

0.38

DJ

0.38

 Frank

0.37

Activations Density 0.000%

No Known Activations

This feature has no known activations.