INDEX

Explanations

attributes or descriptors of places and items that convey exclusivity or trendiness

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

 judiciaire

-0.44

 étoit

-0.44

 juges

-0.42

 lengte

-0.41

 fût

-0.41

 ferons

-0.39

 razon

-0.39

 tallas

-0.38

 someone

-0.38

 prestations

-0.38

POSITIVE LOGITS

Personendaten

0.73

Datuak

0.71

SourceChecksum

0.63

fjspx

0.63

 Савезне

0.62

 OMITBAD

0.59

 ivelany

0.58

findpost

0.57

AccessorTable

0.57

RegistryLite

0.57

Activations Density 0.110%

attributes or descriptors of places and items that convey exclusivity or trendiness

No Comments

No Known Activations