INDEX

Explanations

references to specific car models

oai_token-act-pair · gpt-3.5-turbo

references to specific car models, particularly from Tesla

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

gob

-0.72

 Barton

-0.70

uca

-0.68

 gore

-0.64

 overfl

-0.64

 Greenwald

-0.64

 prejudice

-0.63

english

-0.63

 Bulgar

-0.62

 foreigners

-0.62

POSITIVE LOGITS

 Model

3.89

Model

2.83

 Models

2.56

model

2.34

 model

2.29

 models

2.01

models

1.84

 modeling

1.74

 modelling

1.60

 modeled

1.51

Activations Density 0.014%

references to specific car models

references to specific car models, particularly from Tesla

No Comments

No Known Activations