INDEX

Explanations

terms related to economic concepts and developments

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

atos

-0.68

ornia

-0.65

hari

-0.65

aji

-0.64

bowl

-0.64

 Occupations

-0.64

irable

-0.63

PI

-0.63

 Corpus

-0.63

cribed

-0.63

POSITIVE LOGITS

 nonetheless

1.84

 nevertheless

1.52

etheless

1.44

 still

0.95

 cautioned

0.93

 overshadowed

0.89

 undeniably

0.86

 alas

0.83

 caution

0.82

 undeniable

0.81

Activations Density 4.552%

terms related to economic concepts and developments

No Comments

No Known Activations