INDEX

Explanations

references to specific items or elements in a scientific context

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

 Taktlose

-0.82

featureID

-0.81

OGND

-0.81

+#+

-0.74

MessageTagHelper

-0.70

IUrlHelper

-0.63

WK

-0.63

GenerationType

-0.61

NameInMap

-0.61

ArgsConstructor

-0.61

POSITIVE LOGITS

 stöd

0.60

 støtte

0.51

 oznacza

0.50

 deportivas

0.46

 område

0.46

 zostało

0.46

 Italij

0.45

 hjælp

0.45

 novia

0.45

 cuyo

0.45

Activations Density 0.092%

references to specific items or elements in a scientific context

No Comments

No Known Activations