INDEX

Explanations

phrases that indicate authorship or attribution in texts

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

reporting attribution by name

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

 lenker

-0.66

 autorytatywna

-0.59

insegna

-0.52

blurRadius

-0.52

 FontFamily

-0.51

 Vikipedi

-0.51

 InputDecoration

-0.50

bootstrapcdn

-0.50



-0.49

 AssemblyVersion

-0.49

POSITIVE LOGITS

 Redacción

0.44

by

0.44

 reporter

0.39

By

0.37

://

0.37

By

0.36

 Evans

0.35

 journalist

0.35

 reporters

0.35

(!$

0.34

Activations Density 0.003%

phrases that indicate authorship or attribution in texts

reporting attribution by name

No Comments

No Known Activations