INDEX

Explanations

special characters and specific formatting elements in text

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

numbers and symbols

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

 hinweg

-0.33

anstal

-0.33

expiresIn

-0.31

-0.29

filepath

-0.29

ession

-0.28

SFP

-0.28

 poise

-0.27

Demografía

-0.27

ريقة

-0.27

POSITIVE LOGITS

GEBURTSDATUM

0.80

Personendaten

0.68

 surla

0.68

 GenerationType

0.67

 onPostExecute

0.66

Tembelea

0.66

Personensuche

0.63

uxxxx

0.63

webElementXpaths

0.61

fromnode

0.61

Activations Density 0.007%

special characters and specific formatting elements in text

numbers and symbols

No Comments

No Known Activations