INDEX

Explanations

phrases related to containing or encasing things, possibly with a sense of protection or confinement

oai_token-act-pair · gpt-3.5-turbo

words associated with containment or enclosure

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 icons

-0.83

izons

-0.78

 faults

-0.70

 Guides

-0.69

 buttons

-0.68

 logos

-0.67

 symbols

-0.66

 trademarks

-0.66

 Horizons

-0.66

 labels

-0.66

POSITIVE LOGITS

 manner

1.00

 somewhere

0.92

 reminiscent

0.89

 mode

0.88

 resembling

0.87

 limbo

0.82

 overseen

0.81

 booth

0.81

fitted

0.80

 hurry

0.80

Activations Density 0.381%

phrases related to containing or encasing things, possibly with a sense of protection or confinement

words associated with containment or enclosure

No Comments

No Known Activations