INDEX

Explanations

instances of the word "over" indicating excessive amounts or conditions

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

excess or too much

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

nehin

-0.52

otros

-0.45

 lenker

-0.45

})$}

-0.45

DrawerToggle

-0.45

hrad

-0.44

 torchvision

-0.44

 Pardo

-0.43

Angola

-0.42

JvmStatic

-0.42

POSITIVE LOGITS

 Over

0.50

Over

0.43

StructEnd

0.42

 split

0.41

split

0.41

 over

0.41

 OVER

0.40

OVER

0.39

 blowing

0.38

 Overs

0.38

Activations Density 0.019%

instances of the word "over" indicating excessive amounts or conditions

excess or too much

No Comments

No Known Activations