INDEX

Explanations

numerical data and references in technical documents

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

Numbers within specific formatting (e.g., IDs, codes)

np_acts-logits-general · gemini-2.0-flash

numbers and codes

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.30.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

帖最后由

-0.44

下载附件

-0.39

 ſte

-0.35

 ſta

-0.34

 ſol

-0.31

UnsafeEnabled

-0.31

GEBURTSDATUM

-0.31

الحياه

-0.29

 ſch

-0.29

 Viernes

-0.28

POSITIVE LOGITS

 nonUne

0.59

0.50

úrese

0.50

әрмәләр

0.48

 Bioaccumulative

0.46

 OnDestroy

0.44

ufact

0.44

0.43

URBANA

0.43

Olympia

0.43

Activations Density 1.041%

numerical data and references in technical documents

Numbers within specific formatting (e.g., IDs, codes)

numbers and codes

No Comments

No Known Activations