INDEX

Explanations

actions involving movement and entry

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

security, entry, or access points

np_acts-logits-general · gemini-2.0-flash

gates and doors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.30.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

libft

-0.40

ollectionView

-0.39

Aholisi

-0.38

 Kreise

-0.38

endregion

-0.38

 EClass

-0.35

RectangleBorder

-0.35

 BaseModel

-0.34

onViewCreated

-0.34

身边

-0.34

POSITIVE LOGITS

 door

1.98

 doors

1.81

 entrance

1.80

 gate

1.79

 gates

1.77

 doorway

1.70

 doorways

1.67

door

1.62

 Door

1.62

 entrances

1.60

Activations Density 0.295%

actions involving movement and entry

security, entry, or access points

gates and doors

No Comments

No Known Activations