INDEX

Explanations

terms and concepts related to workplace safety regulations and practices

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

Warnings, safety measures, and potential hazards

np_acts-logits-general · gemini-2.0-flash

safety and hazard warnings

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Einfluß

-0.42

 rijk

-0.40

 vorgenommen

-0.40

 nadzieję

-0.38

 experiências

-0.38

 umowy

-0.37

experiences

-0.36

 Paglinawan

-0.36

*/}

-0.35

Selfer

-0.35

POSITIVE LOGITS

 safety

0.77

 Safety

0.71

 unsafe

0.67

Safety

0.62

 SAFETY

0.61

safety

0.61

 NSCoder

0.59

unsafe

0.57

的安全

0.54

 safe

0.52

Activations Density 0.238%

terms and concepts related to workplace safety regulations and practices

Warnings, safety measures, and potential hazards

safety and hazard warnings

No Comments

No Known Activations