INDEX

Explanations

financial terms related to income or taxes

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

worms

-0.75

loo

-0.72

IGHTS

-0.71

choes

-0.70

gins

-0.67

matic

-0.67

glers

-0.66

ACTION

-0.66

init

-0.66

iannopoulos

-0.65

POSITIVE LOGITS

 household

0.95

 Income

0.87

 Household

0.86

 income

0.85

 incomes

0.75

 respondent

0.72

wealth

0.70

 median

0.69

ered

0.69

 salary

0.69

Activations Density 0.019%

financial terms related to income or taxes

No Comments

No Known Activations