INDEX

Explanations

the word "other" used as a comparative or differentiating term

oai_token-act-pair · gpt-3.5-turbo

references to the word "other"

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

 nicely

-0.64

 legalized

-0.64

 Revival

-0.63

levard

-0.60

 fever

-0.59

 rosters

-0.58

 congrat

-0.56

 capacity

-0.56

 vertically

-0.56

 knees

-0.55

POSITIVE LOGITS

other

4.41

OTHER

1.70

otherapy

1.51

 OTHER

1.49

Other

1.48

Others

1.32

 other

1.31

another

1.23

 Other

1.22

 Others

1.22

Activations Density 0.011%

the word "other" used as a comparative or differentiating term

references to the word "other"

No Comments

No Known Activations