INDEX

Explanations

expressions of gratitude

oai_token-act-pair · gpt-3.5-turbo

expressions of gratitude

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 projecting

-0.73

 Inferno

-0.72

 Loft

-0.70

Osc

-0.70

 deviation

-0.67

 diver

-0.65

 Dise

-0.64

 Ultron

-0.64

 Hurricanes

-0.63

arian

-0.63

POSITIVE LOGITS

gements

0.97

gments

0.92

ifully

0.88

bye

0.83

giving

0.82

 thank

0.82

 Thank

0.82

 acknowled

0.81

Thank

0.78

bles

0.78

Activations Density 0.011%

expressions of gratitude

expressions of gratitude

No Comments

No Known Activations