INDEX

Explanations

quantities and attributes of items or concepts

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 يتيمه

-0.85

 tartalomajánló

-0.82

+#+#

-0.73

 صوتيه

-0.72

 препратки

-0.71

Personendaten

-0.68

Tikang

-0.66

SourceChecksum

-0.65

matchCondition

-0.65

########.

-0.64

POSITIVE LOGITS

is

0.32

def

0.32

tagext

0.28

 incorporado

0.27

 incorporated

0.27

fristig

0.26

 upor

0.26

</strong>

0.26

 partenariat

0.25

</b>

0.25

Activations Density 0.277%

quantities and attributes of items or concepts

No Comments

No Known Activations