INDEX

Explanations

health-related terms and phrases, including health care policies, insurance, outcomes, and concerns

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Reloaded

-0.82

Duo

-0.80

coni

-0.79

 Helpful

-0.77

xual

-0.68

 Darkness

-0.66

chin

-0.66

 Phant

-0.66

 Downing

-0.66

 Gemini

-0.65

POSITIVE LOGITS

 care

1.26

care

1.24

 Care

1.09

span

1.00

Care

0.99

iest

0.95

hazard

0.94

 insurance

0.93

aceutical

0.86

 practition

0.84

Activations Density 0.427%

health-related terms and phrases, including health care policies, insurance, outcomes, and concerns

No Comments

No Known Activations