INDEX

Explanations

references to specific orders or classifications

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

order of

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

-0.45

ifilm

-0.43

German

-0.41

 Brem

-0.40

 بتوان

-0.40

Haf

-0.39

Maybe

-0.39

 Gamb

-0.37

isAuthenticated

-0.37

 Ceci

-0.36

POSITIVE LOGITS

 Order

1.34

 order

1.18

 getOrder

1.16

Order

1.15

order

1.12

 Orders

1.10

 ORDER

1.09

getOrder

1.02

ORDER

0.99

 orders

0.94

Activations Density 0.049%

references to specific orders or classifications

order of

No Comments

No Known Activations