INDEX

Explanations

punctuation related to conditional phrases and lists

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

lists followed by enumerators

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

 arşivlendi

-0.48

/**

-0.45



-0.39

 للاسماء

-0.38

">—

-0.38

 Grüsse

-0.37

 الرياضيه

-0.37

 nakalista

-0.36

 Reſ

-0.35

rungsseite

-0.34

POSITIVE LOGITS

etc

0.82

etc

0.67

Etc

0.57

Etc

0.50

 these

0.50

ect

0.50

ETC

0.48

等等

0.45

usercontent

0.44

ect

0.44

Activations Density 0.051%

punctuation related to conditional phrases and lists

lists followed by enumerators

No Comments

No Known Activations