INDEX

Explanations

significant textual markers that indicate context or emphasis in a document

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

technical or code terms

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

TagMode

-0.49

NameInMap

-0.48

Билгалдахарш

-0.45

 typelib

-0.45



-0.44

Hochspringen

-0.44

Шаг

-0.44

Rohy

-0.43

 okuyayım

-0.42

POSITIVE LOGITS

fxml

0.48

InitVars

0.44

 <<<<<<<<<<<<<<

0.43

parsedMessage

0.41

 ffilmiau

0.41

ViewFeatures

0.41

 initComponents

0.39

mazoo

0.39

Parabéns

0.38

 للمعارف

0.38

Activations Density 0.051%

significant textual markers that indicate context or emphasis in a document

technical or code terms

No Comments

No Known Activations