INDEX

Explanations

references to laboratories and studios

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

 propOrder

-0.58

 Dutch

-0.57

stick

-0.55

 Hindi

-0.54

 Greek

-0.52

 Crohn

-0.52

Pik

-0.52

 HttpRequest

-0.52

 Welsh

-0.52

BorderLayout

-0.50

POSITIVE LOGITS

 bibliothèque

0.65

 studio

0.64

 studios

0.62

 auditorium

0.60

 showroom

0.59

 bookstore

0.59

 Bookstore

0.58

 Showroom

0.57

 Auditorium

0.56

 oficina

0.56

Activations Density 0.447%

references to laboratories and studios

No Comments

No Known Activations