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

itra

-0.16

Ð½Ð³

-0.15

ohana

-0.14

usher

-0.14

etty

-0.14

/Gate

-0.14

unda

-0.14

orca

-0.13

ITES

-0.13

CBS

-0.13

POSITIVE LOGITS

Os

0.24

 Beat

0.24

 beat

0.23

Beat

0.22

beat

0.22

os

0.21

/os

0.18

Bindable

0.18

 Drawable

0.17

plr

0.17

Activations Density 0.000%

No Known Activations

This feature has no known activations.