INDEX

Explanations

articles and pronouns related to human conditions and experiences

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

a followed by specific words

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

 varandra

-0.35

koop

-0.32

Pří

-0.32

Topology

-0.32

NOM

-0.31

 Raub

-0.31

&__

-0.31

 Prä

-0.31

Prä

-0.30

]%

-0.30

POSITIVE LOGITS

GEBURTSDATUM

0.59

 Administrativna

0.56



0.52

 resourceCulture

0.51

 AssemblyTitle

0.51

 PagesJaunes

0.49

onActivityResult

0.48

TypedDataSet

0.48

 gynhyrchwyd

0.48

__':

0.48

Activations Density 0.011%

articles and pronouns related to human conditions and experiences

a followed by specific words

No Comments

No Known Activations