INDEX

Explanations

symbols, formatting characters, and numerical representations in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Italijanski

-0.62

MENAFN

-0.55

 neglect

-0.53

TabIndex

-0.53

addContainerGap

-0.52

 FRANCISCO

-0.52

khana

-0.49

wikimedia

-0.49

Five

-0.48

 CURIAM

-0.48

POSITIVE LOGITS

 hundred

1.15

 eighty

1.03

hundred

1.01

 seventy

1.00

 ninety

0.99

 Hundred

0.94

 Eighty

0.92

 Seventy

0.92

 sixty

0.92

 HUND

0.89

Activations Density 1.529%

symbols, formatting characters, and numerical representations in the text

No Comments

No Known Activations