INDEX

Explanations

proper nouns referring to characters or notable figures

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

nicknames in quotes

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

Microkernel

-0.54

 autorytatywna

-0.49

rawDesc

-0.49

GEBURTSDATUM

-0.48

SpringBootTest

-0.47

ayer

-0.45

 Classification

-0.45

RTLR

-0.43

__*/

-0.43

 gyhoeddwyd

-0.42

POSITIVE LOGITS

InjectAttribute

0.45

sidemargin

0.41

 onOptions

0.41

 lateinit

0.40

venido

0.39

 Italijanski

0.38

 CURIAM

0.38

URIComponent

0.37

IVEREF

0.37

AnimationsModule

0.36

Activations Density 0.003%

proper nouns referring to characters or notable figures

nicknames in quotes

No Comments

No Known Activations