INDEX

Explanations

references to categories or classifications, particularly involving titles and "of" phrases

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

Follows "of" within specific phrases

np_acts-logits-general · gemini-2.0-flash

of [noun] noun phrase

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

onal

-0.62

 standard

-0.58

 Signalez

-0.56

 Standard

-0.54

rawDesc

-0.54

oughly

-0.53

fully

-0.52

tral

-0.51

nä

-0.51

 STANDARD

-0.51

POSITIVE LOGITS

Personendaten

0.62

ControllerAdvice

0.58

 réglable

0.58

endpush

0.57

 kasarigan

0.57

 disambiguazione

0.57

Erfolge

0.55

...")

0.55

ंदीखरीदारी

0.55

UserScript

0.55

Activations Density 0.253%

references to categories or classifications, particularly involving titles and "of" phrases

Follows "of" within specific phrases

of [noun] noun phrase

No Comments

No Known Activations