INDEX

Explanations

mentions of employment-related terms and phrases

oai_token-act-pair · gpt-3.5-turbo

terms related to employment and labor

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

WAYS

-0.68

 aloud

-0.67

 Grip

-0.65

 brace

-0.63

Downloadha

-0.63

 deaf

-0.62

çīĪ

-0.61

 Roche

-0.61

 wholesale

-0.60

 crawl

-0.60

POSITIVE LOGITS

ees

1.32

Employ

1.24

employ

1.20

ments

1.08

ee

1.05

hardt

1.01

ment

0.99

rentices

0.89

estern

0.87

ements

0.86

Activations Density 0.011%

mentions of employment-related terms and phrases

terms related to employment and labor

No Comments

No Known Activations