INDEX

Explanations

references to legal or governmental entities and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 pinulongan

-0.81

\{\\

-0.75

bibfield

-0.65

batore

-0.63

xase

-0.61

__(/*!

-0.60

Personendaten

-0.59

Tikang

-0.58

Geplaatst

-0.57

 utafitiHapana

-0.57

POSITIVE LOGITS

 Canberra

0.69

 Washington

0.57

Washington

0.54

TAWA

0.54

 Potomac

0.53

WASHINGTON

0.51

 capitale

0.51

washingtonpost

0.51

DC

0.50

 Beijing

0.50

Activations Density 1.033%

references to legal or governmental entities and locations

No Comments

No Known Activations