INDEX

Explanations

references to dragons and associated imagery or structures

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

dragon statue

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

argout

-0.36

 pros

-0.35

 commercial

-0.34

mercial

-0.33

 primary

-0.33

 official

-0.33

 cirurgia

-0.31

 period

-0.31

 officially

-0.30

 ufficiali

-0.30

POSITIVE LOGITS

 HasFactory

0.64

bewerken

0.59

expandindo

0.58

Aktualisiert

0.55

Билгалдахарш

0.55

sidemargin

0.54

 виправивши

0.52

قایناق‌لار

0.52

GEBURTSDATUM

0.52

тьяна

0.50

Activations Density 0.014%

references to dragons and associated imagery or structures

dragon statue

No Comments

No Known Activations