INDEX

Explanations

phrases related to disbelief and disbelief expressions

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

CI

-0.64

rant

-0.63

ItemImage

-0.63

senal

-0.62

 Actions

-0.60

CI

-0.60

 bount

-0.60

 littered

-0.59

estate

-0.59

soDeliveryDate

-0.59

POSITIVE LOGITS

 afford

1.46

 muster

1.07

 imagine

1.03

 conceive

1.00

 cope

0.97

athom

0.95

 feas

0.94

 convince

0.94

 stomach

0.94

 decipher

0.93

Activations Density 0.722%

phrases related to disbelief and disbelief expressions

No Comments

No Known Activations