INDEX

Explanations

phrases related to political discussions and strategies

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

expressions of political concern and discussions about electoral strategies

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

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

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_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 explorers

-0.68

 cafes

-0.67

 lantern

-0.63

 Explorer

-0.62

 canopy

-0.61

DIY

-0.61

 Amsterdam

-0.61

 Warehouse

-0.60

 Scrolls

-0.59

 monastery

-0.58

POSITIVE LOGITS

TRUMP

1.49

Hillary

1.40

Trump

1.29

 TRUMP

1.29

Romney

1.28

Democrats

1.27

DonaldTrump

1.27

Republican

1.27

Donald

1.26

GOP

1.25

Activations Density 1.122%

phrases related to political discussions and strategies

expressions of political concern and discussions about electoral strategies

No Known Activations