INDEX

Explanations

phrases referring to digital content or technology

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

xual

-0.94

IELD

-0.82

ortium

-0.78

 Strauss

-0.76

 Krug

-0.74

ered

-0.74

etta

-0.73

nard

-0.71

pell

-0.71

witz

-0.69

POSITIVE LOGITS

ized

1.00

ization

0.97

Ocean

0.95

isation

0.94

netic

0.92

 footprint

0.88

 currency

0.84

izable

0.83

izing

0.81

 literacy

0.81

Activations Density 0.014%

phrases referring to digital content or technology

No Comments

No Known Activations