INDEX

Explanations

the word "dans" indicating location or position

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

NPY

-0.79

 Oviedo

-0.76

 kayo

-0.67

 varandra

-0.67

ASF

-0.66

 sevilla

-0.66

 Genève

-0.63

 perchè

-0.63

 bezeichneter

-0.63

ayuno

-0.62

POSITIVE LOGITS

 Dans

1.22

Dalam

1.21

Dans

1.20

dans

1.18

 dans

1.18

dalam

1.04

 Dalam

1.02

 DANS

0.98

 trong

0.96

 Trong

0.96

Activations Density 0.002%

the word "dans" indicating location or position

No Comments

No Known Activations

the word "dans" indicating location or position

No Comments

No Known Activations