INDEX

Explanations

references to authoritative sources or citations

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

"According to"

np_acts-logits-general · gemini-2.0-flash

according to

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

remel

-0.56

 fileSize

-0.55

ysuckle

-0.54

 itſelf

-0.52

yyb

-0.52

舺

-0.51

 Vergangenheit

-0.50

IImage

-0.50

kesmas

-0.49

createNewFile

-0.49

POSITIVE LOGITS

 Secondo

0.70

Secondo

0.70

Según

0.68

 Según

0.67

 לפי

0.66

según

0.63

rivit

0.63

Selon

0.61

 Selon

0.58

 según

0.56

Activations Density 0.096%

references to authoritative sources or citations

"According to"

according to

No Comments

No Known Activations