INDEX

Explanations

references to news networks and media organizations

oai_token-act-pair · gpt-3.5-turbo

mentions of various "Network" entities, likely in a media or reporting context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uations

-0.69

sql

-0.69

olkien

-0.69

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

-0.68

 Malfoy

-0.64

ffee

-0.63

 reper

-0.63

 awoken

-0.63

gets

-0.62

âĶĢâĶĢâĶĢâĶĢ

-0.62

POSITIVE LOGITS

 Network

1.00

 Networks

0.99

 Tycoon

0.92

anyahu

0.88

eers

0.85

edIn

0.84

hof

0.83

 Rail

0.83

 Security

0.80

 Advertising

0.79

Activations Density 0.015%

references to news networks and media organizations

mentions of various "Network" entities, likely in a media or reporting context

No Comments

No Known Activations