INDEX

Explanations

terms related to medical reports and case documentation

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

keywords indicating document types

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 CreateTagHelper

-0.52

 ویکی‌پدی

-0.48

AndEndTag

-0.47

ftagPool

-0.47

 للاسماء

-0.45

 oscuridad

-0.42

เอง

-0.42

 própria

-0.42

awtextra

-0.42

endregion

-0.41

POSITIVE LOGITS

 description

0.53

strual

0.51

 action

0.51

iële

0.50

 datab

0.50

 besz

0.50

cabul

0.48

IllegalArgument

0.48

chism

0.48

 directional

0.47

Activations Density 0.081%

terms related to medical reports and case documentation

keywords indicating document types

No Comments

No Known Activations