FP-ANR: A representation format to handle floating-point cancellation at run-time.

David Defour

Browse the full ARITH paper archive.