INDEX

Explanations

dates and associated events mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ktop

-0.57

strip

-0.52

exc

-0.51

ropy

-0.51

DragonMagazine

-0.50

cont

-0.50

culated

-0.50

Cons

-0.48

 Tyrann

-0.48

CAP

-0.48

POSITIVE LOGITS

0.86

0.83

0.82

0.81

0.79

ruary

0.78

0.77

Activations Density 0.062%

dates and associated events mentioned in the text

No Comments

No Known Activations