INDEX

Explanations

names of individuals and their roles or occupations

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

Ö¼

-0.68

pher

-0.65

 appar

-0.64

 awoken

-0.64

 plurality

-0.63

 perception

-0.63

forced

-0.62

intention

-0.62

Marginal

-0.61

 footing

-0.60

POSITIVE LOGITS

 Email

1.15

 subscribe

1.12

 Subscribe

1.10

 LINK

1.02

www

1.01

Subscribe

0.98

<|endoftext|>

0.97

 Updates

0.97

 Downloads

0.95

 Join

0.93

Activations Density 4.399%

names of individuals and their roles or occupations

No Comments

No Known Activations