INDEX

Explanations

references to scams and fraudulent behavior

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

scams or fraudulent activity

np_acts-logits-general · gemini-2.0-flash

fraudulent calls and websites

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

ju

-0.30

 coverage

-0.26

 edge

-0.26

-0.24

כות

-0.23

 cool

-0.23

 function

-0.23

 nghiệm

-0.23

нки

-0.23

 تغ

-0.22

POSITIVE LOGITS

 scammers

1.02

 scams

0.98

 scam

0.94

 scammer

0.91

 Scam

0.91

Scam

0.90

 impostor

0.89

 fraudulent

0.87

 المعيارى

0.85

 kasarigan

0.84

Activations Density 0.299%

references to scams and fraudulent behavior

scams or fraudulent activity

fraudulent calls and websites

No Comments

No Known Activations