INDEX

Explanations

locations such as streets, countries, and buildings

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

lopp

-0.55

pmwiki

-0.52

ITNESS

-0.51

)?

-0.51

pired

-0.48

asher

-0.48

Balt

-0.47

?).

-0.47

\":

-0.47

ĨĴ

-0.45

POSITIVE LOGITS

 because

0.85

 indefinitely

0.84

 unless

0.76

 amid

0.76

 until

0.75

 without

0.74

 altogether

0.73

via

0.72

 while

0.72

 irrespective

0.72

Activations Density 0.635%

locations such as streets, countries, and buildings

No Comments

No Known Activations