INDEX

Explanations

significant numerical values related to dates or statistical data

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

years

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



-0.38

getDoctrine

-0.38

Personensuche

-0.36

تقاوى

-0.36

WebElementEntity

-0.35

 ModelRenderer

-0.34

Parcelize

-0.33

 nakalista

-0.33

 poderá

-0.33

Voci

-0.32

POSITIVE LOGITS

 Verſ

0.61

DockStyle

0.58

BibitemShut

0.57

IContainer

0.55

 Audiodateien

0.53

cables

0.52

citenamefont

0.52

cable

0.52

 zwiſchen

0.52

 brutto

0.52

Activations Density 0.106%

significant numerical values related to dates or statistical data

years

No Comments

No Known Activations