INDEX

Explanations

references to royalty

oai_token-act-pair · gpt-3.5-turbo

references to royalty

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

 Loaded

-0.81

 Extend

-0.80

onica

-0.79

Ñı

-0.78

ulhu

-0.75

LER

-0.74

geist

-0.73

 Humanity

-0.73

orem

-0.70

 Canaver

-0.69

POSITIVE LOGITS

 decree

1.00

 palace

0.97

pin

0.96

 pard

0.95

 princes

0.92

 decre

0.89

 prince

0.85

 coffers

0.84

 royal

0.80

 pardon

0.79

Activations Density 0.011%

references to royalty

references to royalty

No Comments

No Known Activations