INDEX

Explanations

comparative adjectives describing health or quality

oai_token-act-pair · gpt-3.5-turbo

phrases related to health assessments and ratings

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

iton

-0.77

ande

-0.77

adium

-0.74

idian

-0.72

idon

-0.72

 weap

-0.70

iversary

-0.66

"]=>

-0.64

*/(

-0.64

NetMessage

-0.63

POSITIVE LOGITS

 mediocre

1.48

 satisfactory

1.34

cellent

1.28

poor

1.25

 excellent

1.23

 poor

1.23

 decent

1.22

Excellent

1.18

 okay

1.17

 dismal

1.16

Activations Density 0.438%

comparative adjectives describing health or quality

phrases related to health assessments and ratings

No Comments

No Known Activations