INDEX

Explanations

mathematical symbols or notations indicating equations or theorems

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

nanomaterials, expressed as

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

 يتيمه

-0.92

 Administrativna

-0.91

 فريبيس

-0.91

 ligiloj

-0.89

StructEnd

-0.88

parsedMessage

-0.88

 мәкал

-0.84

 '\\;'

-0.82

hyrchwyd

-0.82

niſſe

-0.80

POSITIVE LOGITS

Gew

0.33

isSelected

0.30

 marked

0.30

 शुरू

0.28

0.27

 iniz

0.25

 initially

0.25

Setting

0.24

 immediately

0.24

<eos>

0.24

Activations Density 0.001%

mathematical symbols or notations indicating equations or theorems

nanomaterials, expressed as

No Comments

No Known Activations