INDEX

Explanations

repeated usage of the article "the."

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

technical and foreign words

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

 olmayan

-0.47

and

-0.44

of

-0.44

 arrep

-0.41

-0.37

 poderá

-0.36

aimerais

-0.36

 anzunehmen

-0.36

 pourriez

-0.36

 before

-0.36

POSITIVE LOGITS

 ویکی‌پدیا

0.79

 pinulongan

0.63

 useAppContext

0.61

 BoxFit

0.60

PyExc

0.59

expandindo

0.57

хьтан

0.57

 ddelwed

0.57

 queſta

0.56

 estekak

0.56

Activations Density 0.044%

repeated usage of the article "the."

technical and foreign words

No Comments

No Known Activations