INDEX

Explanations

various forms of the word "embarrass" as well as other words related to confusion and complexities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

 Kebijakan

-0.46

 ocio

-0.46

 Lingkungan

-0.42

 Prieto

-0.41

 Pembangunan

-0.40

 Professor

-0.39

ook

-0.39

Ums

-0.39

RenderAtEndOf

-0.38

 Konkurrenz

-0.37

POSITIVE LOGITS

 ModelExpression

0.59

 avoient

0.52

⟬

0.52

LookAnd

0.50

IsMutable

0.50

 coö

0.50

:✨

0.48

 propa

0.48

 auroit

0.47

 pouvoit

0.47

Activations Density 0.489%

various forms of the word "embarrass" as well as other words related to confusion and complexities

No Comments

No Known Activations