INDEX

Explanations

references to data analysis practices and methodologies

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

code, legal judgments, abuse, logs, entities

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

extAlignment

-0.46

 fortæ

-0.44

 surla

-0.43

 Вікіпе

-0.41

 rhestr

-0.40

舺

-0.35

ClientRect

-0.34

 esternos

-0.34

 Bioaccumulative

-0.33

 fís

-0.32

POSITIVE LOGITS

 ویکی‌پدی

1.07

ьаж

0.85

 noDo

0.81

featureID

0.77

EndGlobalSection

0.75

IndentedString

0.75

msgTypes

0.73

Personensuche

0.73

ंदीखरीदारी

0.71

 Administrativna

0.69

Activations Density 31.039%

references to data analysis practices and methodologies

code, legal judgments, abuse, logs, entities

No Comments

No Known Activations