INDEX

Explanations

references to scientific concepts, protocols, and classifications in technical documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 linkovi

-0.39

-0.36

throw

-0.32

findpost

-0.31

்க

-0.31

 prohibit

-0.31

-0.30

 Dusche

-0.30

We

-0.30

other

-0.30

POSITIVE LOGITS

:✨

1.18

 للاسماء

0.67

 Verſ

0.63

耐久

0.57

ArrowToggle

0.57

 intptr

0.56

rbrakk

0.56

 Seller

0.55

yyr

0.55

 uſed

0.54

Activations Density 0.106%

references to scientific concepts, protocols, and classifications in technical documents

No Comments

No Known Activations