INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

 Townsend

-0.15

.usage

-0.14

peed

-0.14

 Matth

-0.14

ä¿Ŀ

-0.14

Cov

-0.13

vard

-0.13

 Tavern

-0.13

Ø¨ÙĬØ¹

-0.13

 magg

-0.13

POSITIVE LOGITS

LES

0.17

accent

0.15

eldon

0.15

okable

0.15

iane

0.14

eref

0.14

atform

0.14

endor

0.14

appoint

0.14

 sclerosis

0.14

Activations Density 0.000%

No Known Activations

This feature has no known activations.