INDEX

Explanations

mathematical formatting and notation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.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

pshire

-0.68

ViewInit

-0.59

 gubern

-0.58

AsUp

-0.55

cillors

-0.54

 lighthouse

-0.53

 Irland

-0.53

 hearsay

-0.52

 BIBSYS

-0.52

 sceptre

-0.51

POSITIVE LOGITS

enumi

0.90

 defaultstate

0.85

tagHelperRunner

0.65

 StatefulWidget

0.65

 TextAlign

0.64

DotNetBar

0.63

stateProvider

0.60

DataAnnotations

0.56

AxisAlignment

0.56

わけ

0.55

Activations Density 0.087%

mathematical formatting and notation

No Comments

No Known Activations