INDEX

Explanations

references to time frames such as months, years, and hours

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

odder

-0.71

Inside

-0.70

Room

-0.64

 Luck

-0.63

Behind

-0.62

 Ahead

-0.62

enf

-0.61

minster

-0.60

elight

-0.60

IGH

-0.60

POSITIVE LOGITS

 millennia

1.00

 period

0.93

 centuries

0.89

 periods

0.88

 decades

0.88

 span

0.86

 generations

0.82

 hundred

0.81

whelming

0.81

 weekend

0.80

Activations Density 0.124%

references to time frames such as months, years, and hours

No Comments

No Known Activations