INDEX

Explanations

references to "WW" or "World Wrestling," indicating a focus on the wrestling industry or related content

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

acks

-0.17

ussions

-0.15

lation

-0.15

ional

-0.15

Ã´t

-0.14

lev

-0.14

yg

-0.14

uden

-0.14

esis

-0.14

ean

-0.13

POSITIVE LOGITS

idor

0.21

WW

0.18

(World

0.17

 Norton

0.17

WW

0.17

ombat

0.15

war

0.15

ãĥ¼ãĥŃ

0.15

ww

0.15

ç¢

0.14

Activations Density 0.010%

references to "WW" or "World Wrestling," indicating a focus on the wrestling industry or related content

No Comments

No Known Activations