INDEX

Explanations

terms related to rights, advocacy, and ethics

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

||||

-0.87

lihood

-0.76

Interstitial

-0.75

ificate

-0.71

 ç¥ŀ

-0.70

arantine

-0.69

onement

-0.69

yss

-0.68

Footnote

-0.66

achment

-0.65

POSITIVE LOGITS

 experts

1.71

 analysts

1.53

 expert

1.48

 specialists

1.44

 analyst

1.41

 campaigners

1.38

 researcher

1.36

 advocates

1.35

 observers

1.32

 activists

1.30

Activations Density 0.285%

terms related to rights, advocacy, and ethics

No Comments

No Known Activations