INDEX

Explanations

phrases related to selling or offering goods for sale

oai_token-act-pair · gpt-3.5-turbo

references to items being sold or available for purchase

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 specificity

-0.63

involved

-0.59

 Staff

-0.57

 Attacks

-0.56

 energies

-0.55

 retribution

-0.54

mit

-0.53

Exit

-0.52

mu

-0.52

eru

-0.51

POSITIVE LOGITS

 separately

1.19

 individually

1.08

 nationwide

1.04

 anonymously

1.03

 cheaply

1.01

 online

1.01

 nationally

1.01

 internationally

0.99

 exclusively

0.99

 worldwide

0.98

Activations Density 0.211%

phrases related to selling or offering goods for sale

references to items being sold or available for purchase

No Comments

No Known Activations