INDEX

Explanations

questions or statements starting with the word "Asked."

oai_token-act-pair · gpt-3.5-turbo

instances of questions or inquiries

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

 quality

-0.66

 suite

-0.65

 destruction

-0.62

 civilization

-0.62

 effect

-0.61

 effects

-0.61

 comp

-0.61

 consolidation

-0.60

mer

-0.59

 impact

-0.58

POSITIVE LOGITS

Asked

3.49

 Asked

2.92

Regarding

1.61

Said

1.40

Refer

1.32

answered

1.31

Ask

1.30

Questions

1.28

Speaking

1.27

 asked

1.25

Activations Density 0.013%

questions or statements starting with the word "Asked."

instances of questions or inquiries

No Comments

No Known Activations