INDEX

Explanations

instances of authority or law enforcement figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 cook

-0.15

anko

-0.15

cook

-0.14

 mails

-0.14

holm

-0.14

 actionPerformed

-0.14

 visited

-0.13

tingham

-0.13

 visits

-0.13

ãĥ¼ãĥ¬

-0.13

POSITIVE LOGITS

 approached

0.30

 approach

0.29

 Approach

0.28

appro

0.27

 approaches

0.26

 Ð¿Ð¾Ð´ÑħÐ¾Ð´

0.25

Appro

0.24

_appro

0.24

 approaching

0.23

 Appro

0.23

Activations Density 0.496%

instances of authority or law enforcement figures

No Comments

No Known Activations