INDEX

Explanations

references to images or photos, particularly those involving personal content

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_131k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.33.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

RUnlock

-0.49

 transfieras

-0.42

 poveznice

-0.40

 pinulongan

-0.40

tvguidetime

-0.40

hability

-0.39

djangoproject

-0.39

>--}}

-0.38

tagHelperRunner

-0.38

VersionUID

-0.37

POSITIVE LOGITS

 smiling

0.75

 yourself

0.72

 wnętr

0.70

 herself

0.69

 scenes

0.68

 landscapes

0.67

 interiors

0.66

 himself

0.64

 interior

0.63

 sunsets

0.62

Activations Density 0.037%

references to images or photos, particularly those involving personal content

No Comments

No Known Activations