INDEX

Explanations

keywords related to financial crimes, specifically money laundering

oai_token-act-pair · gpt-3.5-turbo

references to financial activities, particularly those related to money laundering

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

agonist

-0.67

 Beef

-0.66

 Roof

-0.63

ettings

-0.63

 Historic

-0.62

VD

-0.62

 Aven

-0.62

SPA

-0.61

COMPLE

-0.60

 abdom

-0.60

POSITIVE LOGITS

 laundering

1.59

bags

1.23

making

1.05

 laund

1.04

maker

0.91

flow

0.91

 lenders

0.91

 chang

0.88

bag

0.88

 lender

0.87

Activations Density 0.053%

keywords related to financial crimes, specifically money laundering

references to financial activities, particularly those related to money laundering

No Comments

No Known Activations