INDEX

Explanations

references to locations or nationalities associated with specific individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

 Palembang

-0.33

jupiter

-0.32

 Thessal

-0.31

 deque

-0.30

 mitään

-0.30

 Johor

-0.30

 Czechoslovak

-0.29

Quién

-0.29

ędziesz

-0.28

 sèche

-0.28

POSITIVE LOGITS

 Kurdish

0.83

Kur

0.75

 Kurdistan

0.73

Kur

0.72

 Kurds

0.71

kur

0.71

SourceChecksum

0.70

 المعيارى

0.67

:✨

0.64

featureID

0.63

Activations Density 0.148%

references to locations or nationalities associated with specific individuals

No Comments

No Known Activations