INDEX

Explanations

phrases related to brightness or standing out

oai_token-act-pair · gpt-3.5-turbo

instances of the word "shine" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iet

-0.77

ilateral

-0.71

 Experimental

-0.70

alloc

-0.69

 cannibal

-0.68

 Terms

-0.66

 Termin

-0.65

ort

-0.64

 Site

-0.63

 Hort

-0.63

POSITIVE LOGITS

 shine

4.03

 shines

2.78

 shone

2.56

 shining

2.28

 Shine

2.12

shine

2.09

 illuminate

1.37

 brill

1.34

 spotlight

1.29

elight

1.24

Activations Density 0.017%

phrases related to brightness or standing out

instances of the word "shine" and its variations

No Comments

No Known Activations