Source: cirosantilli/cool-data-embedded-in-the-bitcoin-blockchain/bitcoin-burn-addresses-unveiling-the-permanent-losses-and-their-underlying-causes

= Bitcoin Burn Addresses: Unveiling the Permanent Losses and Their Underlying Causes
{title2=2025-03-18}

https://arxiv.org/pdf/2503.14057 

By Mohamed el Khatib and Arnaud Legout.

Both autors were at <Inria Centre at Université Côte d'Azur>, Mohamed the intern and Arnaud the Inria researcher employee.

Cool, this method could reveal novel <P2FKH> images:
> We propose a methodology to automatically detect burn addresses. We manually classified
208,656 addresses suspected to be burn addresses because they have a low Shannon entropy
> Our model identified 7,905 true burn addresses from a pool of 1,283,997,050 addresses with only 1,767 false positive.
Unfortunately their method might not be well suited for finding images, later on:
> Storing images for fun and posterity. We did not observe plain text messages encoded in Bech32 addresses. As our methodology is designed to identify burn addresses with a human-readable structure or easily identifiable patterns, we are not supposed to detect images encoded in burn addresses.

Data for their results can be found at:

* https://github.com/cirosantilli/bitcoin-inscription-indexer#low-entropy-addresses-khatib-legout-csv
* https://github.com/cirosantilli/bitcoin-inscription-indexer/blob/master/data-manual/low-entropy-addresses-khatib-legout.csv