INDEX

Explanations

numerical values related to statistics or quantities

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

brance

-0.57

GoldMagikarp

-0.57

 Cutter

-0.56

 recognizing

-0.56

 totality

-0.56

 recognition

-0.56

 continual

-0.54

 prominence

-0.53

actionDate

-0.53

 pitched

-0.51

POSITIVE LOGITS

1.50

1.49

1.48

1.46

1.45

1.44

Activations Density 0.136%

numerical values related to statistics or quantities

No Comments

No Known Activations