INDEX

Explanations

names and terms related to North Korea

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

toe

-0.73

 msec

-0.72

sburgh

-0.71

 Normandy

-0.70

bos

-0.69

 COVER

-0.69

 Delaware

-0.68

 Gibbs

-0.67

 Phillies

-0.67

 mileage

-0.67

POSITIVE LOGITS

jin

1.43

 Yuan

1.37

jing

1.31

Zhu

1.25

jiang

1.24

 Jing

1.23

wei

1.23

 Xiao

1.18

yu

1.17

Tai

1.15

Activations Density 3.337%

names and terms related to North Korea

No Comments

No Known Activations