INDEX

Explanations

references to visits, visits to specific places or people

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

bra

-0.75

dp

-0.73

rity

-0.72

corn

-0.67

reg

-0.65

rama

-0.65

alysis

-0.64

claimer

-0.64

percent

-0.63

frac

-0.62

POSITIVE LOGITS

 Disneyland

0.88

 museums

0.83

 abroad

0.81

 exhibitions

0.76

 town

0.76

 Tucson

0.76

 Shang

0.72

 beaches

0.72

 Tours

0.72

 Mongolia

0.71

Activations Density 0.038%

references to visits, visits to specific places or people

No Comments

No Known Activations