INDEX

Explanations

phrases related to evaluating or critiquing something

oai_token-act-pair · gpt-3.5-turbo

sentences that conclude thoughts or ideas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 involuntary

-0.80

 enrol

-0.79

 invitation

-0.74

 volunteered

-0.73

 volunteer

-0.73

 mailed

-0.72

 volunte

-0.72

 disbel

-0.71

 occupation

-0.70

 detention

-0.70

POSITIVE LOGITS

 Whereas

1.65

 Whilst

1.41

 Granted

1.38

 Firstly

1.35

 Nonetheless

1.33

 Therefore

1.32

 Essentially

1.32

 Specifically

1.32

 Regardless

1.29

 Nevertheless

1.28

Activations Density 0.526%

phrases related to evaluating or critiquing something

sentences that conclude thoughts or ideas

No Comments

No Known Activations