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

cak

-0.17

íķĻêµĲ

-0.17

bis

-0.17

_TERMIN

-0.16

belt

-0.16

bpp

-0.16

edReader

-0.16

 resident

-0.15

ees

-0.15

edImage

-0.15

POSITIVE LOGITS

ith

0.34

mond

0.30

gar

0.29

monds

0.28

ible

0.28

vard

0.27

mund

0.27

monton

0.27

inburgh

0.27

ificio

0.26

Activations Density 0.016%

No Comments

No Known Activations