INDEX

Explanations

mentions of educational content and materials

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

educational + related nouns

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

ApiModelProperty

-0.75

 Geben

-0.71

ImageContext

-0.70

 aikaa

-0.69

 varandra

-0.68

 Gelände

-0.65

 Houſe

-0.65

 början

-0.64

 Gegenteil

-0.64

 hırka

-0.64

POSITIVE LOGITS

neurial

0.73

Educational

0.72

 Educational

0.68

educational

0.64

 educational

0.62

 EDUCATIONAL

0.62

変な

0.54

religious

0.54

ductory

0.54

pherical

0.54

Activations Density 0.006%

mentions of educational content and materials

educational + related nouns

No Comments

No Known Activations