INDEX

Explanations

negative or critical comments directed towards politicians or political parties

oai_token-act-pair · gpt-3.5-turbo

instances of strong opinions or criticisms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

 recognised

-0.76

ĸļ

-0.74

 wearable

-0.72

 arte

-0.72

 explorers

-0.68

 Guinness

-0.68

 Grimoire

-0.67

stice

-0.67

 maritime

-0.65

 photography

-0.65

POSITIVE LOGITS

Democrats

1.59

Republicans

1.53

GOP

1.52

 Dems

1.51

Republican

1.43

 Republicans

1.42

 Democrats

1.39

Hillary

1.39

TRUMP

1.38

 Kasich

1.36

Activations Density 0.963%

negative or critical comments directed towards politicians or political parties

instances of strong opinions or criticisms

No Comments

No Known Activations