INDEX

Explanations

words related to authority and control

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

 Kelly

-0.16

 Trophy

-0.15

ç¦ıåĪ©

-0.14

endoza

-0.14

utures

-0.14

Kelly

-0.14

 Pickup

-0.14

à¸±à¸Ļà¸Ļ

-0.14

.osgi

-0.14

ÏģÎ±Î²

-0.14

POSITIVE LOGITS

cz

0.20

sez

0.15

caster

0.15

ceph

0.15

ÐµÐ½ÑĮ

0.15

oref

0.14

rolling

0.14

ampo

0.14

atrix

0.14

rolled

0.14

Activations Density 0.002%

words related to authority and control

No Comments

No Known Activations