INDEX

Explanations

measurements and dimensions

oai_token-act-pair · gpt-3.5-turbo

dimensions and measurements related to objects

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

xit

-0.69

 privatization

-0.69

 refres

-0.69

 contingency

-0.69

 deferred

-0.68

 denial

-0.67

rek

-0.65

issance

-0.65

osph

-0.64

 rewriting

-0.64

POSITIVE LOGITS

 diameter

1.48

 width

1.43

Width

1.33

 centimeters

1.32

 circumference

1.31

 inches

1.23

 Width

1.23

 length

1.21

 height

1.18

 thickness

1.17

Activations Density 0.221%

measurements and dimensions

dimensions and measurements related to objects

No Comments

No Known Activations