INDEX

Explanations

sequences of characters or identifiers that may represent coded or structured data

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

identifiers like UUIDs

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

 jScrollPane

-0.53

 noDo

-0.52

 Lijst

-0.51

 lenker

-0.50

 <=",

-0.50

 سكانية

-0.47

afficheront

-0.47

 useAppContext

-0.47

NewReader

-0.47

ंदीखरीदारी

-0.47

POSITIVE LOGITS

bcryptjs

0.40

aarrggbb

0.40

IntOverflow

0.37

 kasarigan

0.37

addPreferredGap

0.35

 foregoing

0.34

equila

0.33

marginRight

0.33

 piuttosto

0.32

]")]

0.29

Activations Density 0.013%

sequences of characters or identifiers that may represent coded or structured data

identifiers like UUIDs

No Comments

No Known Activations