INDEX

Explanations

instances of professional behavior and etiquette

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

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

OrCreate

-0.15

anki

-0.14

ì§ľ

-0.14

 Bout

-0.13

olla

-0.13

vui

-0.13

ahas

-0.13

aira

-0.13

ÑĥÑĢÐ¸

-0.13

toolbox

-0.13

POSITIVE LOGITS

 etiquette

0.32

 manners

0.29

 etiqu

0.28

 protocol

0.28

 conduct

0.28

 courtesy

0.28

 decor

0.26

 respect

0.26

 comport

0.26

 Courtesy

0.26

Activations Density 0.272%

instances of professional behavior and etiquette

No Comments

No Known Activations