INDEX

Explanations

references to reading articles or reviews

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 generaciones

-0.35

 Architekten

-0.35

 Architekt

-0.34

katapos

-0.33

Hozzáférés

-0.33

 gerekti

-0.33

 Könige

-0.33

 hijas

-0.31

 Komponist

-0.31

 McKenna

-0.31

POSITIVE LOGITS

 resourceCulture

0.79

 nahilalakip

0.73

 disambiguazione

0.71

 ExecuteAsync

0.69

writeFieldEnd

0.68

 autorytatywna

0.67

StructEnd

0.67

LayoutStyle

0.66

SharedDtor

0.65

addCriterion

0.64

Activations Density 0.017%

references to reading articles or reviews

No Comments

No Known Activations