INDEX

Explanations

terms associated with awards and honors in the context of media and entertainment

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

Followed by Canadian media outlets/names

np_acts-logits-general · gemini-2.0-flash

Canadian news and media

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asInstanceOf

-0.78

 Etats

-0.61

 correctes

-0.60

 stockage

-0.59

 dispositif

-0.57

 démoc

-0.57

Defense

-0.57

defn

-0.57

 flavors

-0.57

 forcément

-0.56

POSITIVE LOGITS

 Canadian

1.43

 Canada

1.38

 Canadians

1.36

 RCMP

1.36

 Toronto

1.32

 Ontario

1.31

Canadian

1.31

 Winnipeg

1.28

Canada

1.27

Toronto

1.23

Activations Density 0.406%

terms associated with awards and honors in the context of media and entertainment

Followed by Canadian media outlets/names

Canadian news and media

No Comments

No Known Activations