INDEX

Explanations

various punctuation marks and symbols

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

java/kotlin code libraries and annotations

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

 CascadeType

-0.48

ulemon

-0.46

➯

-0.45

 betweenstory

-0.44

hasMoreElements

-0.43

 wireType

-0.43

廈

-0.41

clerView

-0.41

ImageContext

-0.40

 Theſe

-0.39

POSITIVE LOGITS

 للمعارف

0.57

ьаж

0.56

InjectAttribute

0.46

 TestBed

0.45

springframework

0.44

حياته

0.44

Autoritní

0.43

fjspx

0.43

typeorm

0.41

 Савезне

0.40

Activations Density 0.000%

various punctuation marks and symbols

java/kotlin code libraries and annotations

No Comments

No Known Activations