INDEX

Explanations

titles or mentions of royal figures

oai_token-act-pair · gpt-3.5-turbo

names and titles associated with royalty and nobility

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

ocre

-0.83

inarily

-0.80

 Helpful

-0.78

alore

-0.74

ceptive

-0.74

ILY

-0.71

resso

-0.71

ictive

-0.70

aneous

-0.70

ilater

-0.70

POSITIVE LOGITS

BSD

0.82

 throne

0.76

çİĭ

0.76

jong

0.75

XVI

0.75

XIV

0.73

 Lumpur

0.72

 Chiefs

0.72

 Anniversary

0.71

Zed

0.70

Activations Density 0.206%

titles or mentions of royal figures

names and titles associated with royalty and nobility

No Comments

No Known Activations