INDEX

Explanations

complex medical terms and references related to studies and safety assessments

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

report, technique, continue, practice

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

saraba

-0.56

WebVitals

-0.43

errHandler

-0.43

 propOrder

-0.43

----</

-0.43

 intptr

-0.42

Portály

-0.42

 okuyayım

-0.40

ornos

-0.40

 ſtand

-0.40

POSITIVE LOGITS



0.42

şört

0.41

lood

0.40

 Administrativna

0.39

 facie

0.39

oč

0.38

 femininas

0.38

loed

0.38

TOBER

0.37

 fört

0.37

Activations Density 0.158%

complex medical terms and references related to studies and safety assessments

report, technique, continue, practice

No Comments

No Known Activations