INDEX

Explanations

references to abandoned or empty locations and buildings

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

abandoned or empty places

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

Personendaten

-0.38

 demanding

-0.36

 rushed

-0.34

 demand

-0.33

rrggbb

-0.33

Hauptartikel

-0.32

 Krakowie

-0.31

 relax

-0.31

 relaxing

-0.30

 trasmissione

-0.30

POSITIVE LOGITS

 Biôgrafia

0.65

 abandoned

0.64

istoitu

0.63

abandoned

0.63

 ब्रेकडाउन

0.60

 Abandoned

0.58

 abandonment

0.56

 desolate

0.56

adaptiveStyles

0.55

 deserted

0.54

Activations Density 0.019%

references to abandoned or empty locations and buildings

abandoned or empty places

No Comments

No Known Activations