INDEX

Explanations

words related to direction or alignment

oai_token-act-pair · gpt-3.5-turbo

terms related to orientation, positioning, and alignment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

soDeliveryDate

-0.92

enegger

-0.89

isSpecialOrderable

-0.85

ilk

-0.81

asper

-0.78

bour

-0.76

adows

-0.75

hire

-0.75

ells

-0.73

otte

-0.72

POSITIVE LOGITS

eering

1.23

 orientation

1.00

 shift

0.83

 axis

0.82

 shif

0.82

irection

0.81

ational

0.78

 direction

0.75

 toward

0.74

 symmetry

0.74

Activations Density 0.099%

words related to direction or alignment

terms related to orientation, positioning, and alignment

No Comments

No Known Activations