INDEX

Explanations

words related to work and employment

oai_token-act-pair · gpt-3.5-turbo

references to labor and the workforce

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

irs

-0.84

forth

-0.74

 Flavoring

-0.74

 Robertson

-0.72

kef

-0.71

 Seym

-0.70

Dob

-0.68

Dak

-0.68

Ack

-0.64

Nib

-0.64

POSITIVE LOGITS

 labor

1.02

 labour

0.84

 unions

0.84

 stopp

0.81

Labor

0.80

 laborers

0.80

abor

0.80

Union

0.77

union

0.76

ayers

0.76

Activations Density 0.011%

words related to work and employment

references to labor and the workforce

No Comments

No Known Activations