INDEX

Explanations

words and phrases indicating the presence of research findings or conclusions

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

German words and concepts

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

 ModelExpression

-0.85

 snippetHide

-0.68

########.

-0.65

AsUp

-0.64

 يتيمه

-0.64

تقاوى

-0.62

 الرياضيه

-0.60

-0.59

HtmlAttribute

-0.57

XmlAccessType

-0.57

POSITIVE LOGITS

package

0.36

Teilen

0.32

 depositphotos

0.30

ⓘ

0.29

Portale

0.28

 Berufung

0.28

 waarmee

0.28

 Bestimmung

0.27

ELINES

0.27

Grüße

0.27

Activations Density 0.062%

words and phrases indicating the presence of research findings or conclusions

German words and concepts

No Comments

No Known Activations