A witness-indistinguishable proof (WIP) is a variant of a zero-knowledge proof for languages in NP.
In a typical zero-knowledge proof of a statement, the prover will use a witness for the statement as input to the protocol, and the verifier will learn nothing other than the truth of the statement.
In a WIP, this zero-knowledge condition is weakened, and the only guarantee is that the verifier will not be able to distinguish between provers that use different witnesses.
Witness-indistinguishable proof systems were first introduced by Feige and Shamir.
[1] Unlike zero-knowledge proofs, they remain secure when multiple proofs are being performed in parallel.