INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

Fay

-0.16

 PureComponent

-0.16

-0.15

 late

-0.15

co

-0.15

adio

-0.14

grid

-0.13

 moda

-0.13

aid

-0.13

POSITIVE LOGITS

erin

0.16

ulong

0.15

ioso

0.15

egin

0.15

FieldType

0.14

ìĥģ

0.14

 nhÃ¡nh

0.14

 ná»ķi

0.14

ÏĢÎ¹

0.14

Activations Density 0.098%

No Comments

No Known Activations