INDEX

Explanations

patterns of high numerical values that denote importance or relevance in data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

AsUp

-0.93

 lenker

-0.81

 Bourgoin

-0.80

 CreateTagHelper

-0.78

fromnode

-0.76

adaptiveStyles

-0.75

createCanvas

-0.74

 мәкал

-0.73

Datuak

-0.70

にほんブログ村

-0.67

POSITIVE LOGITS

1.74

1.61

1.51

1.39

1.37

1.35

1.34

1.33

1.25

1.23

Activations Density 0.118%

patterns of high numerical values that denote importance or relevance in data

No Comments

No Known Activations