INDEX

Explanations

mentions or references to the word "Trump."

oai_token-act-pair · gpt-3.5-turbo

mentions of Donald Trump

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iencies

-0.69

ASA

-0.66

NAD

-0.64

KDE

-0.63

Eh

-0.62

 Emacs

-0.60

 Rena

-0.60

 Desmond

-0.59

 Kali

-0.59

 bred

-0.58

POSITIVE LOGITS

care

1.05

 supporters

1.01

 Tower

1.01

Jr

1.00

TRUMP

0.95

 supporter

0.93

eter

0.93

Jr

0.93

eting

0.92

 Supporters

0.91

Activations Density 0.064%

mentions or references to the word "Trump."

mentions of Donald Trump

No Comments

No Known Activations