INDEX

Explanations

elements related to data queries and filtering criteria

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

criteria, type, genre

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

numerusform

-0.50

IsContent

-0.50

AnimationsModule

-0.49

 CascadeType

-0.49

endphp

-0.49

AsUp

-0.48

ReusableCell

-0.48

CPM

-0.47

bindingNavigator

-0.46

POSITIVE LOGITS

 kasarigan

0.49

 Verlags

0.48

 Innenstadt

0.40

 Eisenbahn

0.40

Források

0.40

 Riggs

0.40

 Nadel

0.40

 Öffnungszeiten

0.40

 Spielen

0.39

 Câmara

0.38

Activations Density 0.009%

elements related to data queries and filtering criteria

criteria, type, genre

No Comments

No Known Activations