INDEX

Explanations

questions and topics related to various discussions and debates

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

rites

-0.81

urses

-0.73

nice

-0.70

 fixme

-0.68

ellow

-0.65

ensions

-0.64

inters

-0.64

 agre

-0.61

ufact

-0.61

gian

-0.60

POSITIVE LOGITS

naires

1.18

 unanswered

1.15

 arises

0.95

naire

0.95

 whether

0.88

 questions

0.86

 mark

0.83

 arise

0.83

 posed

0.82

 plag

0.82

Activations Density 0.059%

questions and topics related to various discussions and debates

No Comments

No Known Activations