INDEX

Explanations

references to the concept of being able or capable

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

able followed by action or attribute

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

Negative Logits

 Durchmesser

-0.42

 these

-0.41

 jugement

-0.40

 Marchand

-0.39

toCharArray

-0.38

 Fong

-0.38

 Konzentration

-0.38

 granito

-0.38

 queima

-0.38

 Entfernung

-0.37

POSITIVE LOGITS

Able

1.26

 Able

1.24

 able

0.85

Abel

0.83

 Abel

0.78

Willing

0.75

Ability

0.73

 capable

0.72

 Ability

0.69

 Willing

0.68

Activations Density 0.014%

references to the concept of being able or capable

able followed by action or attribute

No Known Activations