INDEX

Explanations

words related to medical conditions and assistive devices

oai_token-act-pair · gpt-3.5-turbo

references to wheelchairs and related themes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Gal

-0.74

ial

-0.67

Inst

-0.66

 Firm

-0.65

ought

-0.65

 Markets

-0.64

tell

-0.63

Met

-0.63

 Mutual

-0.61

aught

-0.61

POSITIVE LOGITS

 wheelchair

4.14

chairs

1.50

 disability

1.44

ADA

1.41

 Disability

1.27

 disabilities

1.26

 disabled

1.23

 handic

1.20

 cane

1.20

 chair

1.18

Activations Density 0.012%

words related to medical conditions and assistive devices

references to wheelchairs and related themes

No Comments

No Known Activations