INDEX

Explanations

references to dog-related behaviors and training

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 continúas

-0.37

ByVersion

-0.36

makeText

-0.36

rrggbb

-0.36

 يتيمه

-0.35

Hochspringen

-0.33

LEncoder

-0.31

abbix

-0.31

ischof

-0.30

娥

-0.30

POSITIVE LOGITS

dog

3.13

 dogs

2.95

Dog

2.67

 Dogs

2.63

Dog

2.58

Dogs

2.50

 canine

2.50

dog

2.50

DOG

2.41

dogs

2.39

Activations Density 0.396%

references to dog-related behaviors and training

No Known Activations