INDEX

Explanations

phrases related to personal or political interests

oai_token-act-pair · gpt-3.5-turbo

references to self-serving behaviors and interests in decision-making

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

drop

-0.87

 Trace

-0.76

DragonMagazine

-0.74

etheless

-0.74

xit

-0.71

wayne

-0.69

anmar

-0.69

dra

-0.69

ername

-0.68

ori

-0.67

POSITIVE LOGITS

 interests

1.55

interest

1.34

 profit

1.30

 convenience

1.24

 agendas

1.22

 profits

1.16

 selfish

1.16

 gratification

1.12

 interest

1.12

 greed

1.12

Activations Density 0.519%

phrases related to personal or political interests

references to self-serving behaviors and interests in decision-making

No Comments

No Known Activations