INDEX

Explanations

various levels of whitespace and formatting in the text

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

sizeCache

-0.85

 تضيفلها

-0.73

ngdoc

-0.65

 JpaRepository

-0.65

InSection

-0.59

AnchorStyles

-0.57

 Мексичка

-0.57

 AssemblyTitle

-0.57

 oprot

-0.57

تقاوى

-0.56

POSITIVE LOGITS

 bikini

0.50

 Bikini

0.46

letics

0.42

Hea

0.42

Play

0.42

Execute

0.42

’

0.41

Kaw

0.41

Logo

0.41

Wildcard

0.40

Activations Density 0.061%

various levels of whitespace and formatting in the text

No Comments

No Known Activations