INDEX

Explanations

news-related keywords and phrases

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

knife

-0.60

 Sparrow

-0.58

sole

-0.57

otin

-0.56

lly

-0.55

bos

-0.55

 congr

-0.54

Sap

-0.54

gran

-0.53

 asteroids

-0.53

POSITIVE LOGITS

 Videos

0.95

 Thumbnails

0.81

 VIDEOS

0.74

 WATCHED

0.72

 Video

0.66

iques

0.65

});

0.65

Video

0.64

 videos

0.63

 Lessons

0.60

Activations Density 0.011%

news-related keywords and phrases

No Comments

No Known Activations