INDEX

Explanations

specific formatting or structured data representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

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

olsom

-0.52

 Insufficient

-0.49

ThroughAttribute

-0.48

ırı

-0.48



-0.48

 Curi

-0.48

PointSize

-0.47

 Signs

-0.47

LayoutStyle

-0.46

 fédé

-0.46

POSITIVE LOGITS

 otomatig

0.83

 مرئيه

0.78

 дописавши

0.69

Personendaten

0.68

NewUrlParser

0.63

ruptcy

0.61

EndGlobalSection

0.60

expandindo

0.59

GOTREF

0.59

Jîn

0.54

Activations Density 0.013%

specific formatting or structured data representations

No Comments

No Known Activations