INDEX

Explanations

references to dice and related terms in gaming contexts

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

dice and rolls

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

AddTagHelper

-0.64

Życiorys

-0.62

SourceChecksum

-0.61

 Majefty

-0.59

 ujednoznacz

-0.58

findpost

-0.57

 invokingState

-0.55

存于互联网档案馆

-0.54

 arşivlendi

-0.53

 للاسماء

-0.53

POSITIVE LOGITS

 dice

0.95

Dice

0.93

 Dice

0.91

dice

0.78

 DICE

0.68

🎲

0.65

 dices

0.57

 Ludo

0.55

骰

0.54

 diced

0.50

Activations Density 0.002%

references to dice and related terms in gaming contexts

dice and rolls

No Comments

No Known Activations