INDEX

Explanations

phrases related to interior design and social interactions happening in a specific setting

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

national

-0.86

 specialization

-0.78

 Historically

-0.78

abase

-0.78

 historically

-0.77

itutional

-0.76

 Founding

-0.75

 nominees

-0.75

conservancy

-0.72

 domestically

-0.71

POSITIVE LOGITS

 startled

1.37

 murm

1.32

 nervously

1.29

 grinning

1.27

 grin

1.26

 stares

1.25

 gigg

1.24

 whispered

1.22

 softly

1.22

 smir

1.21

Activations Density 6.414%

phrases related to interior design and social interactions happening in a specific setting

No Comments

No Known Activations