INDEX

Explanations

superlative descriptors related to global prominence and ranking

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

world's superlatives

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

SequentialGroup

-0.59

tvguidetime

-0.56

 CURIAM

-0.55

+#+

-0.52

 Allociné

-0.51

 ModelExpression

-0.49

 själva

-0.48

 OnTriggerEnter

-0.46

 Insee

-0.45

 NgModule

-0.45

POSITIVE LOGITS

 World

0.66

世界上

0.66

world

0.65

 world

0.64

World

0.63

 WORLD

0.61

 Worldwide

0.58

 earth

0.57

世上

0.56

 Earth

0.56

Activations Density 0.006%

superlative descriptors related to global prominence and ranking

world's superlatives

No Comments

No Known Activations