INDEX

Explanations

the presence of the term "et" in the document

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

et al. citations

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

 columnName

-0.44

 destro

-0.44

able

-0.43

 inger

-0.43

 something

-0.42

braun

-0.41

 Something

-0.41

 للاسماء

-0.41

something

-0.41

own

-0.40

POSITIVE LOGITS

et

0.95

Et

0.77

 ligiloj

0.68

et

0.63

Et

0.60

Cet

0.59

 cetera

0.57

ET

0.55

etches

0.55

 Infórmanos

0.54

Activations Density 0.021%

the presence of the term "et" in the document

et al. citations

No Comments

No Known Activations