INDEX

Explanations

words related to consequences or instructions/action items in various scenarios

oai_token-act-pair · gpt-3.5-turbo

discussions around consequence or societal impact

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

License

-0.76

Had

-0.72

MpServer

-0.69

REDACTED

-0.63

 looph

-0.63

tained

-0.63

hound

-0.60

ãĤ´ãĥ³

-0.58

PokÃ©

-0.57

acher

-0.57

POSITIVE LOGITS

 invariably

1.27

 usually

1.21

 typically

1.09

 inevitably

1.07

usually

1.05

often

0.93

 often

0.92

 tends

0.91

ometimes

0.87

Enlarge

0.86

Activations Density 0.412%

words related to consequences or instructions/action items in various scenarios

discussions around consequence or societal impact

No Comments

No Known Activations