INDEX

Explanations

phrases related to service or serving

oai_token-act-pair · gpt-3.5-turbo

words related to service and servitude

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

 prone

-0.64

 Kubrick

-0.64

 fertile

-0.63

 heavier

-0.61

 curly

-0.60

 dense

-0.60

 heaviest

-0.59

 Pound

-0.59

razil

-0.59

 invasive

-0.59

POSITIVE LOGITS

serv

1.22

iced

1.22

maid

1.00

icer

0.97

icing

0.96

iments

0.95

 serv

0.95

Serv

0.90

 Serv

0.88

itor

0.87

Activations Density 0.016%

phrases related to service or serving

words related to service and servitude

No Comments

No Known Activations