INDEX

Explanations

the name "Donald Trump" in various contexts

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 @ 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

pole

-0.81

 dots

-0.77

 foss

-0.68

erb

-0.67

duct

-0.65

esville

-0.64

 notebooks

-0.64

shapeshifter

-0.63

papers

-0.63

kers

-0.62

POSITIVE LOGITS

 Jinping

0.76

terness

0.75

Bid

0.73

 Abrams

0.70

thora

0.68

 Marshall

0.67

 Doyle

0.66

°

0.66

ª

0.65

 Griffin

0.64

Activations Density 0.116%

the name "Donald Trump" in various contexts

mentions of Donald Trump

No Comments

No Known Activations