INDEX

Explanations

references to programming issues and technical challenges

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

expandindo

-0.67

^(@)

-0.64

 Bomba

-0.63

byes

-0.63

 poffible

-0.60

 GoFundMe

-0.58

 RSSSF

-0.57

 douce

-0.56

AnchorTagHelper

-0.56

 pleaf

-0.56

POSITIVE LOGITS

GEBURTSDATUM

0.80

tagHelperRunner

0.73

OGND

0.65

 nahilalakip

0.63

 мәкал

0.61

хьтан

0.61

Aholisi

0.60

 فريبيس

0.60

saraba

0.57

NameInMap

0.55

Activations Density 1.314%

references to programming issues and technical challenges

No Comments

No Known Activations