INDEX

Explanations

proper nouns and phrases related to individuals or entities known for specific characteristics or actions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating expertise or recognition associated with individuals or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

soever

-0.92

UME

-0.74

ixon

-0.73

MRI

-0.73

ieties

-0.70

soDeliveryDate

-0.70

atar

-0.67

ISE

-0.66

OTAL

-0.66

DOI

-0.66

POSITIVE LOGITS

 producing

1.23

 being

1.20

 delivering

1.18

 having

1.16

 possessing

1.15

 creating

1.14

 portraying

1.10

 invent

1.09

 embod

1.08

 providing

1.08

Activations Density 0.128%

proper nouns and phrases related to individuals or entities known for specific characteristics or actions

phrases indicating expertise or recognition associated with individuals or groups

No Comments

No Known Activations