INDEX

Explanations

warnings and cautionary advice related to health or safety

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

caution or warning

np_acts-logits-general · gemini-2.0-flash

unsafe or risky situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 تضيفلها

-0.66

anggil

-0.61

الإنجليزية

-0.58

 sederhana

-0.55

tvb

-0.53

ējās

-0.53

numerusform

-0.52

ادگی

-0.51

følgelig

-0.51

SuspendLayout

-0.50

POSITIVE LOGITS

 risky

0.77

 potentially

0.77

 flammable

0.77

 unsupervised

0.75

 exposed

0.75

 unprotected

0.75

potentially

0.74

 excessively

0.74

unsafe

0.74

 слишком

0.73

Activations Density 0.425%

warnings and cautionary advice related to health or safety

caution or warning

unsafe or risky situations

No Comments

No Known Activations