INDEX

Explanations

nouns related to numbers or quantities

oai_token-act-pair · gpt-3.5-turbo

references to significant societal issues or events impacting people

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

sweet

-0.76

ascus

-0.75

Anth

-0.75

Arc

-0.73

Sweet

-0.73

anamo

-0.73

isite

-0.71

grade

-0.70

 Auburn

-0.69

 spar

-0.69

POSITIVE LOGITS

people

1.82

 people

1.81

 persons

1.65

 person

1.62

 People

1.60

People

1.59

 individuals

1.47

 Person

1.46

 PEOPLE

1.37

person

1.37

Activations Density 0.289%

nouns related to numbers or quantities

references to significant societal issues or events impacting people

No Comments

No Known Activations