INDEX

Explanations

terms related to publication or distribution conditions and statuses

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

published or not

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

makeConstraints

-0.63

 विश्वसनीयता

-0.50

liferay

-0.46

 trust

-0.45

 Gruß

-0.44

extrême

-0.44

 sereia

-0.44

 alfombra

-0.44

 сър

-0.44

 aDecoder

-0.43

POSITIVE LOGITS

istoitu

0.52

 propOrder

0.40

 gyhoeddwyd

0.39

AsUp

0.39

Cyfarwyddwr

0.37

BrowserModule

0.36

 podjela

0.36

adaptiveStyles

0.36

Попис

0.35

ſelf

0.35

Activations Density 0.001%

terms related to publication or distribution conditions and statuses

published or not

No Comments

No Known Activations