INDEX

Explanations

numerical data and specific formatting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 19-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_19/width_16k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.19.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

 wrap

-1.05

 wrapping

-0.97

 Wrap

-0.96

 plastic

-0.93

bag

-0.92

 foil

-0.90

wrap

-0.90

 Foil

-0.88

 bags

-0.88

 cello

-0.87

POSITIVE LOGITS

 autorytatywna

0.68

 linkovi

0.57

LookAnd

0.52

Autoritní

0.50

WebElementEntity

0.49

 المعيارى

0.48

 Soto

0.48

GOTREF

0.47

Com

0.47

 ligiloj

0.47

Activations Density 0.717%

numerical data and specific formatting

No Comments

No Known Activations