INDEX

Explanations

instances of negation or refusal

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

didSet

-0.62

endpush

-0.56

Etymology

-0.50

zá

-0.46

__':

-0.45

 Winn

-0.43

&=&\

-0.43

/>\

-0.43

acquisto

-0.42

bindParam

-0.42

POSITIVE LOGITS

Hentet

0.68

óngase

0.68

ibatis

0.66

windowFixed

0.66

 vuitton

0.65

 defaultstate

0.65

 nahilalakip

0.64

RenderAtEndOf

0.64

tagHelperRunner

0.63

WebVitals

0.62

Activations Density 0.125%

instances of negation or refusal

No Comments

No Known Activations