INDEX

Explanations

the word "very" used to emphasize intensity

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

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.10.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

chanind/openwebtext-gemma

Features

16,384

Data Type

float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

ctigges/openwebtext-gemma-1024-cl

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<bos>

-2.57

ⓧ

-1.01

/**

-0.81

///**

-0.78

/***

-0.77

 endow

-0.77

 intersper

-0.73

 defray

-0.72

//*/

-0.67

<?

-0.66

POSITIVE LOGITS

 VERY

0.97

VERY

0.95

 very

0.95

Very

0.94

 Very

0.93

very

0.91

 marea

0.88

 Portugu

0.80

 Décembre

0.78

 sappi

0.78

Activations Density 0.091%

the word "very" used to emphasize intensity

No Comments

No Known Activations