INDEX

Explanations

phrases related to personal stories and experiences

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

 percentages

-0.72

 ILCS

-0.69

 concessions

-0.68

IST

-0.65

ancing

-0.64

essen

-0.63

 scenarios

-0.62

*/(

-0.61

ance

-0.61

 doses

-0.61

POSITIVE LOGITS

 hers

1.59

 ours

1.56

 yours

1.46

 mine

1.41

 theirs

1.32

 Mine

1.01

 sorts

0.89

mire

0.80

Mine

0.76

irlf

0.70

Activations Density 0.098%

phrases related to personal stories and experiences

No Comments

No Known Activations