INDEX

Explanations

references to restaurants and their operations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

Follows numbers or certain words ("group", "eyes")

np_acts-logits-general · gemini-2.0-flash

franchise or chain locations describing quantities

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 propOrder

-0.69

 autorytatywna

-0.66

Hentet

-0.66

MemoryWarning

-0.65

 Familienname

-0.62

Personendaten

-0.61

bollah

-0.61

apimachinery

-0.60

balleur

-0.59

PreAuthorize

-0.59

POSITIVE LOGITS

 franchise

1.61

 franch

1.38

 franchises

1.37

 Franchise

1.31

 franchi

1.30

franchise

1.30

Franchise

1.27

 chain

1.17

 Franch

1.10

 franqu

1.09

Activations Density 0.223%

references to restaurants and their operations

Follows numbers or certain words ("group", "eyes")

franchise or chain locations describing quantities

No Known Activations