INDEX

Explanations

phrases related to boundaries, especially when they are crossed or set

oai_token-act-pair · gpt-3.5-turbo

phrases related to crossing moral or ethical boundaries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 headquartered

-0.69

ãĥīãĥ©

-0.62

ivery

-0.59

 Plans

-0.59

umped

-0.59

 Tycoon

-0.58

 roundup

-0.58

NECT

-0.57

UMP

-0.57

fab

-0.57

POSITIVE LOGITS

otomy

0.83

 separating

0.80

 between

0.74

 boundaries

0.69

wcs

0.66

 limits

0.66

izons

0.66

posts

0.64

geon

0.64

ourgeois

0.64

Activations Density 0.286%

phrases related to boundaries, especially when they are crossed or set

phrases related to crossing moral or ethical boundaries

No Comments

No Known Activations