INDEX

Explanations

attributes related to quality, clarity, and aesthetics in descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

AttributeSet

-0.46

transQ

-0.44

ARROLLO

-0.44

httphttps

-0.44

chilas

-0.44

 snippetHide

-0.42

 renegoti

-0.42

ScopeManager

-0.41

principalColumn

-0.41

 Erfahrungen

-0.40

POSITIVE LOGITS

 cleanliness

0.88

 clean

0.80

 immaculate

0.79

 spotless

0.75

 perfection

0.75

 pristine

0.73

 purity

0.71

clean

0.67

 Purity

0.67

 Clean

0.66

Activations Density 0.502%

attributes related to quality, clarity, and aesthetics in descriptions

No Comments

No Known Activations