INDEX

Explanations

various forms of the word "unique" and related concepts that imply distinctiveness or particularity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 AppBar

-0.50

 insuffisamment

-0.46

 ioutil

-0.45

 salaire

-0.43

 BoxDecoration

-0.43

 Fernsehen

-0.42

 françaises

-0.41

 Infór

-0.40

 Reglamento

-0.39

ọi

-0.39

POSITIVE LOGITS

 unique

1.09

 UNIQUE

1.04

 uniqueness

1.03

 uniquely

1.02

 Unique

1.02

unique

0.97

Unique

0.97

UNIQUE

0.94

 uniques

0.93

 unik

0.88

Activations Density 0.133%

various forms of the word "unique" and related concepts that imply distinctiveness or particularity

No Comments

No Known Activations