INDEX

Explanations

references to areas and measurements in various contexts

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

area fraction

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

HomeAsUpEnabled

-0.47

Nach

-0.43

 informée

-0.43

ChangePassword

-0.42

зня

-0.41

="#"><

-0.41

switch

-0.40

 raccomand

-0.40

 tergantung

-0.40

correspond

-0.40

POSITIVE LOGITS

 Fläche

0.82

面積

0.81

 Flächen

0.79

面积

0.79

Area

0.77

 Area

0.77

 area

0.75

 acreage

0.73

 AREA

0.67

 área

0.63

Activations Density 0.040%

references to areas and measurements in various contexts

area fraction

No Comments

No Known Activations