INDEX

Explanations

references to gardens and gardening-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 nonUne

-0.38

 executive

-0.38

hObject

-0.38

首

-0.37

 Executive

-0.37

 bubble

-0.36

msup

-0.34

 Brau

-0.33

кну

-0.32

schrift

-0.32

POSITIVE LOGITS

 garden

2.38

 yard

2.28

garden

2.06

Yard

2.02

Garden

2.00

 Garden

1.98

 backyard

1.96

 Yard

1.95

yard

1.88

 GARDEN

1.88

Activations Density 0.055%

references to gardens and gardening-related terms

No Comments

No Known Activations