INDEX

Explanations

programming syntax elements and symbols

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

code structure keywords

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

 lenker

-0.63

transQ

-0.61

TargetException

-0.61

niſſe

-0.60

нгред

-0.55

osd

-0.54

 beſte

-0.54

 Anſ

-0.54

ſammen

-0.52

 Verſ

-0.52

POSITIVE LOGITS

лтемелер

0.46

RefNanny

0.36

 inalámbrico

0.35

MemoryWarning

0.35

 enfin

0.34

 communs

0.34

 seraient

0.33

 تضيفلها

0.33

 parvenir

0.33

 Kecil

0.32

Activations Density 0.003%

programming syntax elements and symbols

code structure keywords

No Comments

No Known Activations