INDEX

Explanations

numerical data patterns and phone numbers

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

phone numbers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

ViewInit

-0.57

posedge

-0.54

AccessFile

-0.54

URLException

-0.53

 desmotivaciones

-0.53

存于互联网档案馆

-0.52

دانشنامهٔ

-0.50

Życiorys

-0.49

XmlAccessType

-0.49

NameInMap

-0.48

POSITIVE LOGITS

phone

0.48

 phone

0.48

 phones

0.46

mobile

0.45

 mobile

0.42

Mobile

0.42

 doors

0.41

alama

0.41

0.40

vices

0.40

Activations Density 0.024%

numerical data patterns and phone numbers

phone numbers

No Comments

No Known Activations