INDEX

Explanations

references to positions of authority or management

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

uien

-0.16

ascimento

-0.16

reau

-0.15

á»ķ

-0.15

roken

-0.15

assium

-0.15

á»«a

-0.15

owego

-0.14

ÏĢÎµÎ¹

-0.14

PLICATION

-0.14

POSITIVE LOGITS

 leadership

0.43

 roles

0.37

 leading

0.32

 Leadership

0.32

 capacities

0.30

 lead

0.30

 role

0.30

leaders

0.29

 dual

0.29

 positions

0.28

Activations Density 0.129%

references to positions of authority or management

No Comments

No Known Activations

references to positions of authority or management

No Comments

No Known Activations