INDEX

Explanations

references to various types of tours

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 nød

-0.52

Personendaten

-0.46

 esternos

-0.45

 kvinder

-0.43

 altså

-0.42

 alcuna

-0.42

 Accommodation

-0.41

 šķ

-0.41

 skolen

-0.40

 abstrato

-0.40

POSITIVE LOGITS

 tour

1.38

 toured

1.28

tour

1.27

 touring

1.20

 tours

1.16

Tour

1.16

 Tour

1.12

 TOUR

1.08

 Touring

0.98

tours

0.98

Activations Density 0.102%

references to various types of tours

No Comments

No Known Activations