INDEX

Explanations

the word "much" and its variations, indicating an emphasis on quantity or extent

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

imax

-0.18

oad

-0.17

asser

-0.16

 little

-0.15

 Little

-0.14

ieren

-0.14

sez

-0.14

DEX

-0.14

prising

-0.14

ictions

-0.14

POSITIVE LOGITS

-needed

0.29

ado

0.28

acho

0.22

acha

0.20

ÑĩÐ¸Ñģ

0.18

/all

0.17

needed

0.17

ãĤĵãģ©

0.15

-expanded

0.15

 obliged

0.15

Activations Density 0.030%

the word "much" and its variations, indicating an emphasis on quantity or extent

No Comments

No Known Activations