INDEX

Explanations

references to washing or cleaning actions

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

cleaning: wash, washing, washed

np_acts-logits-general · gemini-2.0-flash

wash

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

richTextPanel

-0.73

PreferredItem

-0.66

 ResourceManager

-0.57

 CURIAM

-0.54

 propOrder

-0.54

følge

-0.53

 plads

-0.49

tahankan

-0.49

spesies

-0.49

Personendaten

-0.48

POSITIVE LOGITS

 wash

1.81

Wash

1.70

 washes

1.69

 washing

1.65

 Wash

1.64

 washed

1.58

wash

1.50

Washing

1.48

 Washing

1.45

washing

1.42

Activations Density 0.134%

references to washing or cleaning actions

cleaning: wash, washing, washed

wash

No Comments

No Known Activations