INDEX

Explanations

mentions of viewers

oai_token-act-pair · gpt-3.5-turbo

references to television viewers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 reorgan

-0.68

 Surgery

-0.66

 Bain

-0.64

agna

-0.62

 contingency

-0.62

 Bicycle

-0.61

 Shotgun

-0.61

 communist

-0.60

ulic

-0.60

 taxing

-0.59

POSITIVE LOGITS

 viewers

1.02

 tuned

0.94

 viewership

0.91

wat

0.83

 viewer

0.81

 Ratings

0.79

 tuning

0.78

 Roku

0.76

 viewing

0.76

ockets

0.75

Activations Density 0.015%

mentions of viewers

references to television viewers

No Comments

No Known Activations