INDEX

Explanations

references to prior reports or studies

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

previous, past, or general works

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

 مرئيه

-0.56

AddTagHelper

-0.52

 Winaray

-0.49

OuterAlt

-0.47

ſammen

-0.45



-0.45

ereo

-0.45

StringWriter

-0.44

 CURIAM

-0.44

 Viss

-0.43

POSITIVE LOGITS

Personensuche

0.48

 defStyle

0.43

 Biôgrafia

0.41

XMLSchema

0.40

StructEnd

0.39

 EconPapers

0.38

 estekak

0.38

endphp

0.37

Cyfarwyddwr

0.37

 вообще

0.36

Activations Density 0.070%

references to prior reports or studies

previous, past, or general works

No Comments

No Known Activations