INDEX

Explanations

negative statements related to inability or restriction

oai_token-act-pair · gpt-3.5-turbo

negations regarding capability or potential

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

 swaps

-0.71

 tweaks

-0.68

 unve

-0.63

announced

-0.63

 acron

-0.62

 Printed

-0.61

 Cosponsors

-0.60

Factory

-0.59

WithNo

-0.59

cloth

-0.57

POSITIVE LOGITS

 succeed

1.36

 survive

1.33

 thrive

1.25

 prosper

1.22

 deserve

1.21

 enjoy

1.12

 flourish

1.08

 qualify

1.08

 earn

1.08

 attain

1.06

Activations Density 0.326%

negative statements related to inability or restriction

negations regarding capability or potential

No Comments

No Known Activations