INDEX

Explanations

references to social media posts

oai_token-act-pair · gpt-3.5-turbo

references to online social media content, specifically posts

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

orporated

-0.76

eria

-0.74

é¾į

-0.71

estro

-0.68

bid

-0.67

geist

-0.66

aphael

-0.65

erion

-0.65

 EntityItem

-0.64

ochet

-0.64

POSITIVE LOGITS

 posts

3.92

posts

2.68

 postings

2.57

 Posts

2.49

 post

1.98

Posts

1.82

 posting

1.75

 tweets

1.74

 blogs

1.66

 articles

1.62

Activations Density 0.007%

references to social media posts

references to online social media content, specifically posts

No Comments

No Known Activations