INDEX

Explanations

references to self-identification and categorical attributes related to data structures or APIs

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

return false

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

GEBURTSDATUM

-0.73

parsedMessage

-0.71

 Мексичка

-0.65

 الحره

-0.63

djangoproject

-0.62

styleType

-0.60

ItemBackground

-0.59

 OMITBAD

-0.58

isContained

-0.57

UIControlState

-0.56

POSITIVE LOGITS

ⓧ

0.43

Thus

0.38

Also

0.37

 Also

0.37

Additionally

0.36

str

0.36

WS

0.35

 Thus

0.35

 Additionally

0.34

 dins

0.34

Activations Density 0.005%

references to self-identification and categorical attributes related to data structures or APIs

return false

No Comments

No Known Activations