INDEX

Explanations

patterns and special characters often associated with formatted text or data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

 AppCompatTheme

-0.46

getJSONObject

-0.44

scribers

-0.43

GEBURTSDATUM

-0.43

 setEmail

-0.43

ategy

-0.42

SharedCtor

-0.42

pended

-0.41

مراجع

-0.41

TextBoxColumn

-0.41

POSITIVE LOGITS

 kaarangay

0.54

 defaultstate

0.52

تقاوى

0.48

Personendaten

0.47

Personensuche

0.44

Spoljašnje

0.43

 canst

0.40

tagHelperRunner

0.39

 <>",

0.39

sizeCache

0.37

Activations Density 0.337%

patterns and special characters often associated with formatted text or data

No Comments

No Known Activations