INDEX

Explanations

references to products, their features, and user experiences related to software or goods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.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

 placing

-0.42

 dedic

-0.38

 putting

-0.35

 designing

-0.32

 devoting

-0.32

 chose

-0.31

 lägga

-0.31

 wür

-0.30

 wrote

-0.30

Put

-0.30

POSITIVE LOGITS

 disappear

0.73

 undergoes

0.73

 undergo

0.71

 existed

0.70

 exist

0.70

 disappeared

0.68

 disappears

0.67

 remain

0.66

 underwent

0.66

Sklici

0.66

Activations Density 2.189%

references to products, their features, and user experiences related to software or goods

No Comments

No Known Activations