INDEX

Explanations

proper nouns, specifically names of individuals and categories related to people

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

 للاسماء

-0.42

expandindo

-0.37

 propOrder

-0.36

iedział

-0.35

 Numerade

-0.35

DIPSETTING

-0.35

хьтан

-0.34

too

-0.33

 cuch

-0.33

الدراسه

-0.32

POSITIVE LOGITS

CloseOperation

0.61

exitRule

0.59

RetentionPolicy

0.54

webElement

0.53

ſelben

0.50

 scrip

0.48

########.

0.48

acamata

0.47

 BoxDecoration

0.47

ſelf

0.46

Activations Density 0.008%

proper nouns, specifically names of individuals and categories related to people

No Comments

No Known Activations