INDEX

Explanations

terms related to numerical data and measurements

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

silently transfer

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

maakt

-0.43

Jeografia

-0.42

 ब्रेकडाउन

-0.42

IndentedString

-0.41

featureID

-0.39

tanleria

-0.39

 intelig

-0.38

 būs

-0.36

 emigrated

-0.36

 råd

-0.35

POSITIVE LOGITS

OGND

0.54

 يتيمه

0.53

ViewImports

0.52

TemporalType

0.52

 للمعارف

0.52

 Вікіпе

0.50

libft

0.48

Alike

0.47

 EconPapers

0.47

">//

0.47

Activations Density 0.000%

terms related to numerical data and measurements

silently transfer

No Comments

No Known Activations