INDEX

Explanations

symbols and markup related to coding or document structure

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

`# define`

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

XmlAccessType

-0.64

躇

-0.52

 lenker

-0.47

getWriter

-0.47

 consort

-0.46

TintMode

-0.45

ComponentName

-0.44

 itſelf

-0.44

leſs

-0.44

 సౌకర్య

-0.43

POSITIVE LOGITS

 surla

0.49

 ostavi

0.41

 مشين

0.39

Einzelnachweise

0.37

 chied

0.36

сылкі

0.35

rawDesc

0.35

Tē

0.34

səhifə

0.33

}`)

0.33

Activations Density 0.001%

symbols and markup related to coding or document structure

`# define`

No Comments

No Known Activations