INDEX

Explanations

references to yoga and related practices

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

yoga, Yogi, yogurt

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Behaup

-0.41

 herinner

-0.40

////////////////

-0.39

]');

-0.39

 bericht

-0.38

 useRef

-0.37

 Gubernur

-0.37

 Meldung

-0.36

Kı

-0.36

 miljoen

-0.36

POSITIVE LOGITS

 yoga

1.43

Yoga

1.40

 Yoga

1.39

yoga

1.29

 YOGA

1.24

ulemon

1.17

yog

0.96

ヨガ

0.85

yog

0.84

Yog

0.84

Activations Density 0.002%

references to yoga and related practices

yoga, Yogi, yogurt

No Comments

No Known Activations