INDEX

Explanations

instances of the word "due," indicating causality or reasons for events or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 drapeau

-0.47

 ſche

-0.44

Copyright

-0.43

 abſ

-0.43

 paſſ

-0.39

 abstrait

-0.39

 amenaza

-0.39

Eſ

-0.38

 preſent

-0.38

lysninger

-0.38

POSITIVE LOGITS

due

1.55

due

1.33

DUE

1.14

Due

1.12

Due

1.10

DUE

0.96

 devido

0.92

 debido

0.88

 owed

0.86

 owing

0.79

Activations Density 0.159%

instances of the word "due," indicating causality or reasons for events or conditions

No Comments

No Known Activations