INDEX

Explanations

references to digital media or technology professionals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

bard

-0.15

fty

-0.15

 GDPR

-0.14

¢åįķ

-0.14

piry

-0.13

 Birth

-0.13

edith

-0.13

ycastle

-0.13

 discrimin

-0.13

 Marriage

-0.13

POSITIVE LOGITS

 Kelly

0.27

 Ramos

0.27

 Officer

0.26

 officer

0.24

Kelly

0.23

 Officers

0.23

 officers

0.23

 unarmed

0.21

 Manuel

0.21

 cops

0.21

Activations Density 0.003%

references to digital media or technology professionals

No Comments

No Known Activations