INDEX

Explanations

phrases related to being the first to do something

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.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

quickShipAvailable

-0.73

ĺħ

-0.66

EStream

-0.64

ernand

-0.63

 Migration

-0.61

ancies

-0.61

 Powers

-0.59

vind

-0.59

ancy

-0.58

Delivery

-0.58

POSITIVE LOGITS

 admit

1.08

 congratulate

1.02

 introduce

0.98

 acknowledge

0.94

 educate

0.90

 realise

0.85

 realize

0.84

ggles

0.83

 recognize

0.82

 demonstrate

0.81

Activations Density 0.061%

phrases related to being the first to do something

No Comments

No Known Activations