INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

isco

-0.16

enheim

-0.16

rello

-0.16

esel

-0.15

ORAGE

-0.14

 Clause

-0.14

abaj

-0.14

steen

-0.14

ghost

-0.14

imos

-0.14

POSITIVE LOGITS

 hair

0.23

 Hair

0.22

Hair

0.20

 hairstyle

0.19

hair

0.19

 hairstyles

0.19

 locks

0.17

pom

0.17

bob

0.17

bob

0.17

Activations Density 0.000%

No Known Activations

This feature has no known activations.