INDEX

Explanations

patterns related to structured data or programming elements

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 Pró

-0.56

 Enth

-0.55

BLO

-0.52

 entire

-0.50

 utafitiHapana

-0.48

্ল

-0.48

 minib

-0.47

LEY

-0.47

CIOS

-0.47

ợp

-0.47

POSITIVE LOGITS

 CreateTagHelper

0.94

TagMode

0.85

AddTagHelper

0.76

])),

0.74

)))),

0.72

</tr>

0.72

])));

0.67

queryInterface

0.67

MigrationBuilder

0.66

WriteTagHelper

0.65

Activations Density 0.286%

patterns related to structured data or programming elements

No Comments

No Known Activations