INDEX

Explanations

the word "Tesla"

oai_token-act-pair · gpt-3.5-turbo

mentions of the company 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

english

-0.73

Mai

-0.64

rising

-0.64

 raiding

-0.64

 Irish

-0.64

 arch

-0.63

lig

-0.63

cgi

-0.62

ioch

-0.61

ill

-0.60

POSITIVE LOGITS

 Tesla

3.89

Tesla

3.56

 Musk

1.89

 Elon

1.87

 SpaceX

1.82

 Porsche

1.57

 Nissan

1.54

 Nikola

1.50

 Volvo

1.48

 Lyft

1.46

Activations Density 0.016%

the word "Tesla"

mentions of the company Tesla

No Comments

No Known Activations