INDEX

Explanations

mentions of people living, working, or settling in specific places

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 actu

-0.66

catentry

-0.63

neum

-0.60

å£«

-0.60

Series

-0.58

otent

-0.58

 psychiatrists

-0.58

apist

-0.57

urses

-0.57

 Decision

-0.57

POSITIVE LOGITS

 nearby

1.36

 elsewhere

1.35

 here

1.23

 downtown

1.23

 therein

1.18

 outdoors

1.16

 inland

1.14

 north

1.13

 south

1.12

 ashore

1.10

Activations Density 0.609%

mentions of people living, working, or settling in specific places

No Comments

No Known Activations