INDEX

Explanations

mentions of the name "Donald Trump"

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

nexus

-0.76

 lineback

-0.69

 arteries

-0.66

Offline

-0.62

Reviewed

-0.61

 microphones

-0.60

NP

-0.60

road

-0.59

 informants

-0.58

modules

-0.58

POSITIVE LOGITS

 Trump

1.13

 Donald

0.95

trump

0.83

swer

0.83

Trump

0.82

 Sterling

0.78

 Enrique

0.76

 TRUMP

0.76

care

0.75

realDonaldTrump

0.75

Activations Density 0.012%

mentions of the name "Donald Trump"

No Comments

No Known Activations