INDEX

Explanations

references to specific educational roles and environments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 ligiloj

-0.49

InjectAttribute

-0.47

 refroidissement

-0.43

 forbindelse

-0.39

 popolare

-0.37

 ceremonia

-0.36

 prawda

-0.36

 živ

-0.36

 Grüsse

-0.36

tagHelperRunner

-0.36

POSITIVE LOGITS

site

0.75

grid

0.65

 pulled

0.63

 Site

0.62

Site

0.59

queryInterface

0.59

 SITE

0.59

Fair

0.59

 Exit

0.58

 Fair

0.57

Activations Density 0.144%

references to specific educational roles and environments

No Comments

No Known Activations