INDEX

Explanations

phrases related to ability or inability to perform actions.

oai_token-act-pair · gpt-3.5-turbo

negations or contractions indicating the absence or unavailability of something

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

 Tags

-0.73

cano

-0.72

 intensify

-0.69

sonian

-0.68

inav

-0.67

kamp

-0.66

olitics

-0.66

ggles

-0.66

Fail

-0.66

ologne

-0.65

POSITIVE LOGITS

 able

1.51

 aware

1.44

 bothered

1.35

 willing

1.29

 consulted

1.22

 afraid

1.22

 interested

1.19

 thrilled

1.19

 satisfied

1.17

 expecting

1.17

Activations Density 0.208%

phrases related to ability or inability to perform actions.

negations or contractions indicating the absence or unavailability of something

No Comments

No Known Activations