INDEX

Explanations

phrases related to responsibilities and duties

oai_token-act-pair · gpt-3.5-turbo

references to entities and their responsibilities or roles

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SPONSORED

-0.89

cipled

-0.83

osponsors

-0.82

quickShipAvailable

-0.78

Cry

-0.78

kat

-0.74

Times

-0.72

 guiActiveUnfocused

-0.71

atility

-0.70

lass

-0.70

POSITIVE LOGITS

 overseeing

1.86

 enforcing

1.60

 administering

1.59

 coordinating

1.48

 ensuring

1.46

 regulating

1.45

 maintaining

1.42

 protecting

1.40

 guarding

1.40

 interpreting

1.39

Activations Density 0.226%

phrases related to responsibilities and duties

references to entities and their responsibilities or roles

No Comments

No Known Activations