INDEX

Explanations

words or phrases related to specific concepts or items in diverse contexts, often including technical or specialized language

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 schermata

-0.53

 informée

-0.49

 وتسجيلات

-0.45

يكب

-0.45

 فريبيس

-0.43

 esposo

-0.41

 crucero

-0.41

 ब्रेकडाउन

-0.40

 simplifié

-0.39

qxd

-0.39

POSITIVE LOGITS

endregion

0.52

PerformLayout

0.52

 mAdapter

0.50

IContainer

0.50



0.49

LabelTagHelper

0.47

 StatelessWidget

0.47

SequentialGroup

0.47

 تكبرها

0.46

addContainerGap

0.44

Activations Density 0.217%

words or phrases related to specific concepts or items in diverse contexts, often including technical or specialized language

No Comments

No Known Activations