INDEX

Explanations

references to various laboratories, including National Laboratories and academic research institutions

oai_token-act-pair · gpt-3.5-turbo

mentions of various laboratories

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

generic

-0.81

atra

-0.80

yright

-0.76

atar

-0.73

loss

-0.72

quer

-0.72

shaw

-0.71

shr

-0.70

mitt

-0.69

corn

-0.69

POSITIVE LOGITS

 Laboratory

0.98

 laboratory

0.95

 Laboratories

0.91

 laboratories

0.89

rador

0.87

 chemist

0.83

 Technician

0.80

opath

0.79

 technician

0.78

 Analy

0.74

Activations Density 0.025%

references to various laboratories, including National Laboratories and academic research institutions

mentions of various laboratories

No Comments

No Known Activations