INDEX

Explanations

phrases related to income, earnings, and tax brackets

oai_token-act-pair · gpt-3.5-turbo

information related to income and earnings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NK

-0.92

CONCLUS

-0.77

canon

-0.77

DragonMagazine

-0.74

Ú

-0.73

OLOG

-0.73

laun

-0.72

rawdownloadcloneembedreportprint

-0.72

 Maul

-0.71

 Tosh

-0.71

POSITIVE LOGITS

 earners

1.15

 incomes

1.14

amacare

1.07

 income

1.05

 retirees

1.02

 paycheck

1.01

 uninsured

1.01

wage

0.96

elfare

0.94

 subsistence

0.92

Activations Density 0.352%

phrases related to income, earnings, and tax brackets

information related to income and earnings

No Comments

No Known Activations