INDEX

Explanations

numerical data and symbols relevant to scientific measurements and calculations

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

time units and scientific notation

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

 المعيارى

-0.88

AndEndTag

-0.73

TagMode

-0.69

 BoxFit

-0.68

TypedDataSet

-0.67

adaptiveStyles

-0.65

Bakgrunnsstoff

-0.64

__':

-0.61

IsMutable

-0.61

__":

-0.60

POSITIVE LOGITS

źni

0.35

Ἁ

0.33

Has

0.32

Kesimpulan

0.32

faßt

0.32

一

0.32

 wobei

0.32

Has

0.31

 Consultez

0.31

lihatkan

0.30

Activations Density 0.108%

numerical data and symbols relevant to scientific measurements and calculations

time units and scientific notation

No Comments

No Known Activations