INDEX

Explanations

phrases related to being compelled or forced to do something

oai_token-act-pair · gpt-3.5-turbo

instances of being coerced or forced into actions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

atur

-0.66

utm

-0.63

 Juda

-0.62

 Niet

-0.61

roit

-0.61

Avg

-0.60

hess

-0.59

 stead

-0.59

rising

-0.59

pling

-0.58

POSITIVE LOGITS

 available

1.07

 unavailable

1.02

 compulsory

0.97

 redundant

0.96

 aware

0.94

 inaccessible

0.91

 obsolete

0.87

 extinct

0.87

available

0.85

 accessible

0.83

Activations Density 0.118%

phrases related to being compelled or forced to do something

instances of being coerced or forced into actions or conditions

No Comments

No Known Activations