INDEX

Explanations

religious and spiritual references, including names of religious figures and quotations related to faith

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 underwater

-0.70

 pricey

-0.65

 guiActiveUnfocused

-0.65

 pric

-0.64

 nearby

-0.62

 silicone

-0.61

Speed

-0.60

 cylinders

-0.59

 competitor

-0.59

 topping

-0.59

POSITIVE LOGITS

 Chomsky

1.06

omsky

1.02

 Zionist

0.95

 Zionism

0.91

 Socialist

0.89

 neocons

0.88

partisan

0.87

 Orwell

0.86

 Trotsky

0.86

 Feminist

0.84

Activations Density 1.581%

religious and spiritual references, including names of religious figures and quotations related to faith

No Comments

No Known Activations