INDEX

Explanations

question marks indicating inquiries or requests for information

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

questions and answers

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

short question headings (FAQ-style interrogative lines like "What...", "Who...", "Why...")

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

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

Geplaatst

-0.63

 estekak

-0.60

httphttps

-0.60

 desmotivaciones

-0.59

évaluateur

-0.58

 transfieras

-0.57

devamını

-0.55

FDRE

-0.53

Rüyada

-0.53

 surla

-0.53

POSITIVE LOGITS

 resourceCulture

0.45

0.42

?-

0.41

?}

0.40

?=

0.40

?/

0.40

Because

0.39

?:

0.38

Yes

0.37

 Because

0.37

Activations Density 0.055%

question marks indicating inquiries or requests for information

questions and answers

short question headings (FAQ-style interrogative lines like "What...", "Who...", "Why...")

No Comments

No Known Activations