INDEX

Explanations

romantic-related words and concepts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases related to romantic themes and relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

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_post

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

Rah

-0.68

ifted

-0.68

 Welch

-0.67

Reviewer

-0.67

grim

-0.67

izen

-0.65

ocumented

-0.64

 Hait

-0.63

 Greenwood

-0.63

uge

-0.63

POSITIVE LOGITS

 relationships

1.00

 monog

0.99

 romance

0.98

 partner

0.96

 adultery

0.96

 fantasies

0.95

 breakup

0.93

 comed

0.93

 couples

0.92

 flirt

0.92

Activations Density 0.180%

romantic-related words and concepts

phrases related to romantic themes and relationships

No Comments

No Known Activations