INDEX

Explanations

references to calculations or comparisons involving mathematical or statistical values

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

 informée

-0.73

RenderAtEndOf

-0.65

 Administrativna

-0.63

 OFDb

-0.53

 שוליים

-0.51

 tartalomajánló

-0.51

 Italijani

-0.50

 ब्रेकडाउन

-0.47

 calendriers

-0.46

GEBURTSDATUM

-0.46

POSITIVE LOGITS

 ſhould

0.43

StateToProps

0.41

 muſt

0.40

 étoit

0.40

 iſt

0.40

 zoude

0.39

 esfuer

0.39

 paſſ

0.39

 ſte

0.38

 Houſe

0.37

Activations Density 0.717%

references to calculations or comparisons involving mathematical or statistical values

No Comments

No Known Activations