INDEX

Explanations

elements related to brand names or prominent companies

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

 للمعارف

-0.47

-0.46

ViewInit

-0.44

 समीक्षाओं

-0.43

 disambiguazione

-0.42

Билгалдахарш

-0.42

 AssemblyVersion

-0.41

]-->

-0.40

__["

-0.40

}))

-0.39

POSITIVE LOGITS

 BoxFit

0.47

fjspx

0.45

Chham

0.43

 gulls

0.42

 argint

0.41

VYMaps

0.41

 aught

0.41

carboxylic

0.40

Cordialement

0.40

 AssemblyCompany

0.39

Activations Density 0.000%

elements related to brand names or prominent companies

No Comments

No Known Activations