INDEX

Explanations

names of individuals and authors

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

names and titles

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

TagMode

-0.89

 ErrIntOverflow

-0.88

 Савезне

-0.83

 الرياضيه

-0.81

хьтан

-0.77

WebElementEntity

-0.76

GEBURTSDATUM

-0.75

 ivelany

-0.75

 kaarangay

-0.75

ngdoc

-0.73

POSITIVE LOGITS

 along

0.37

상

0.33

ress

0.33

for

0.32

Rus

0.31

and

0.31

 giri

0.31

Mu

0.31

0.30

Activations Density 0.004%

names of individuals and authors

names and titles

No Comments

No Known Activations