INDEX

Explanations

repeated phrases and elements within a text

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

scheduling scheduling

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

SharedCtor

-0.58

 propOrder

-0.56

Personendaten

-0.54

 BitmapFactory

-0.48

 للمعارف

-0.47

featureID

-0.47

styleType

-0.47

 useSelector

-0.46

cobra

-0.44

parsedMessage

-0.43

POSITIVE LOGITS

 nahilalakip

0.52

 Futter

0.40

Erreferentziak

0.39

 vôtre

0.39

Hourly

0.39

 schedule

0.38

⎤

0.38

 scheduling

0.38

 Eigentü

0.38

 schedules

0.37

Activations Density 0.004%

repeated phrases and elements within a text

scheduling scheduling

No Comments

No Known Activations