INDEX

Explanations

specific geographical locations and names related to regions or districts

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

place names following specific words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 AssemblyCulture

-0.83

IBarButtonItem

-0.79

rungsseite

-0.78

Mountain

-0.76

 mountaine

-0.74

 Italijani

-0.74

mountain

-0.73

 gunung

-0.73

 فريبيس

-0.72

 MOUNTAIN

-0.72

POSITIVE LOGITS

 Parcelable

0.34

setFlags

0.30

 cotton

0.29

 peanut

0.28

 الرياضيه

0.28

 peanuts

0.27

 Braunschweig

0.27

 Egipto

0.27

 Neutr

0.27

WebServlet

0.26

Activations Density 0.092%

specific geographical locations and names related to regions or districts

place names following specific words

No Comments

No Known Activations