INDEX

Explanations

references to detailed descriptions and discussions around various topics

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

 footing

-0.71

 Reincarnated

-0.67

 extrad

-0.65

 Trog

-0.63

 endeav

-0.61

 estranged

-0.61

 ideal

-0.61

 communications

-0.61

lling

-0.60

 charm

-0.60

POSITIVE LOGITS

ÂłÂłÂłÂł

1.34

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

1.29

ÂłÂłÂł

1.29

ÂłÂłÂłÂłÂłÂłÂłÂł

1.28

Anyway

1.20

Article

1.05

Posted

1.05

ÂłÂł

1.05

posted

0.99

Advertisement

0.97

Activations Density 2.463%

references to detailed descriptions and discussions around various topics

No Comments

No Known Activations