INDEX

Explanations

promotional codes and instructions for discounts or free items

oai_token-act-pair · gpt-3.5-turbo

promotional content related to offers and discounts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Stalin

-0.68

 politic

-0.66

hov

-0.64

SPONSORED

-0.63

 leftists

-0.63

death

-0.63

wat

-0.61

ÄŁ

-0.61

 Reconstruction

-0.60

 Islamists

-0.59

POSITIVE LOGITS

 personalized

0.84

 customized

0.82

 purchase

0.76

 customizable

0.74

 premium

0.71

 customization

0.69

 vendor

0.69

 coupon

0.69

 functionality

0.69

 custom

0.68

Activations Density 2.601%

promotional codes and instructions for discounts or free items

promotional content related to offers and discounts

No Comments

No Known Activations