INDEX

Explanations

words related to automation or autonomous entities

oai_token-act-pair · gpt-3.5-turbo

references to autonomous and autistic-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 trough

-0.71

 Jenna

-0.68

 thirsty

-0.68

 downed

-0.68

WWE

-0.67

 watering

-0.67

KH

-0.66

 outnumbered

-0.66

 Holly

-0.65

 Wendy

-0.65

POSITIVE LOGITS

aut

4.50

auts

2.66

Aut

1.98

AUT

1.86

autical

1.79

Aut

1.59

AUT

1.39

aut

1.31

Autom

1.28

auto

1.22

Activations Density 0.014%

words related to automation or autonomous entities

references to autonomous and autistic-related concepts

No Comments

No Known Activations