INDEX

Explanations

phrases related to ongoing processes or situations

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

insula

-0.75

stice

-0.74

ridges

-0.74

 Stim

-0.72

alties

-0.66

velt

-0.63

upuncture

-0.63

RTX

-0.63

DragonMagazine

-0.62

ongyang

-0.61

POSITIVE LOGITS

birth

1.02

 undecided

0.94

born

0.93

 reeling

0.91

 intact

0.89

 alive

0.84

 debating

0.78

 able

0.76

 waiting

0.75

 awaiting

0.75

Activations Density 0.052%

phrases related to ongoing processes or situations

No Comments

No Known Activations