INDEX

Explanations

phrases indicating clarity or evident understanding

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

clear, apparent, evident

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 moks

-0.47

 mousse

-0.41

 pillows

-0.40

harusnya

-0.40

SOUNDBITE

-0.39

 mourut

-0.38

/*@

-0.37

 velours

-0.37

 galinha

-0.37

matricula

-0.37

POSITIVE LOGITS

 evident

0.87

evident

0.81

obvious

0.75

 CURIAM

0.72

 obvious

0.69

httphttps

0.67

Obvious

0.65

 duidelijk

0.64

 clear

0.61

clearly

0.58

Activations Density 0.023%

phrases indicating clarity or evident understanding

clear, apparent, evident

No Comments

No Known Activations