INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 Stranger

-0.15

anner

-0.15

ibi

-0.14

æĥĳ

-0.14

 asynchronously

-0.14

ksen

-0.14

eniable

-0.13

æĽľ

-0.13

ivr

-0.13

roi

-0.13

POSITIVE LOGITS

natural

0.15

 Natural

0.15

ystal

0.15

 Naturally

0.15

 natural

0.14

ewn

0.14

acro

0.14

nz

0.13

fec

0.13

Pip

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.