INDEX

Explanations

mentions of the concept of humanity

oai_token-act-pair · gpt-3.5-turbo

references to humanity, often in a critical or negative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

urations

-0.77

Bey

-0.72

WARE

-0.69

Sym

-0.69

FU

-0.67

é¾įå

-0.66

abet

-0.66

 excerpts

-0.64

enegger

-0.64

raph

-0.64

POSITIVE LOGITS

ankind

0.94

 beings

0.82

idis

0.73

 humankind

0.71

 civilisation

0.70

 evolves

0.70

arily

0.69

 inhab

0.68

 civilization

0.68

zee

0.68

Activations Density 0.030%

mentions of the concept of humanity

references to humanity, often in a critical or negative context

No Comments

No Known Activations