INDEX

Explanations

references to different versions of media

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

 version

-2.02

 versions

-1.73

 versión

-1.46

 versiones

-1.29

 versão

-1.27

 Versions

-1.26

 versione

-1.23

 versie

-1.22

版本

-1.13

 Version

-1.08

POSITIVE LOGITS

 consultato

0.61

jsxFileName

0.56

 Vikipedi

0.56

 Maiden

0.53

adaptiveStyles

0.52

ViewFeatures

0.50

 locos

0.49

age

0.47

 noDo

0.47

oredCriteria

0.47

Activations Density 0.168%

references to different versions of media

No Comments

No Known Activations