INDEX

Explanations

references to the word "heart" and its various forms and associations

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

الدراسه

-0.49

 Coup

-0.47

Nix

-0.47

BN

-0.47

 Picasso

-0.47

Pax

-0.46

 Mose

-0.46

);?>

-0.44

Tui

-0.44

 suppose

-0.44

POSITIVE LOGITS

Heart

1.27

 Heart

1.15

heart

1.14

 HEART

1.12

HEART

1.09

 heart

1.02

Hearts

1.02

 Hearts

0.96

hearts

0.96

 cœurs

0.96

Activations Density 0.011%

references to the word "heart" and its various forms and associations

No Comments

No Known Activations