INDEX

Explanations

mentions of royalty, specifically related to the term "Queen."

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

Į¨

-0.20

memberOf

-0.17

oso

-0.15

ãĥ³ãĤ¯

-0.15

ody

-0.15

ark

-0.15

aceous

-0.15

xes

-0.15

ylan

-0.14

å¦»

-0.14

POSITIVE LOGITS

stown

0.21

iyet

0.21

ie

0.21

 Elizabeth

0.19

borough

0.19

 Victoria

0.19

pins

0.17

 Anne

0.16

bee

0.16

ibe

0.16

Activations Density 0.006%

mentions of royalty, specifically related to the term "Queen."

No Comments

No Known Activations