INDEX

Explanations

phrases indicating actions related to reading content

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

Tokens preceding ellipses or continuation of text

np_acts-logits-general · gemini-2.0-flash

more content indicators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

__*/

-1.17

__':

-0.85

__(/*!

-0.73

 समीक्षाएं

-0.71

الحياه

-0.69

 indisponible

-0.68

__':

-0.68

بوابة

-0.68

ThemeOverlay

-0.67

رشف

-0.66

POSITIVE LOGITS

<eos>

1.31

↵↵

0.51

<unused60>

0.50

 urethra

0.46

<unused63>

0.45

"]}

0.43

Попис

0.41

brz

0.41

 chitar

0.41

<unused61>

0.40

Activations Density 0.179%

phrases indicating actions related to reading content

Tokens preceding ellipses or continuation of text

more content indicators

No Comments

No Known Activations