INDEX

Explanations

terms related to classification and categorization of strategies, conditions, or cases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

GraphicsUnit

-0.67

forRoot

-0.60

曖昧さ回避

-0.54

制

-0.54

tahankan

-0.53

featureID

-0.51

 disponibilités

-0.51

 архивлан

-0.48

ButterKnife

-0.46

Personensuche

-0.45

POSITIVE LOGITS

 belongs

1.96

 belong

1.93

 belonging

1.66

 belonged

1.62

belong

1.61

 pertenece

1.39

 classified

1.39

belongs

1.38

 category

1.35

 Belong

1.35

Activations Density 0.848%

terms related to classification and categorization of strategies, conditions, or cases

No Comments

No Known Activations