Manuscript Oai276-A@1 · Oai276-A.tex · section · 215068af

Hypothesis ledger

Stage 4: change one hypothesis at a time › Hypothesis ledger · lines 1710–1758

Change or proposed inference	What actually changes	Disposition
Change or proposed inference	What actually changes	Disposition
Replace classical messages by an unknown quantum input	The coding criterion and target output change.	No quantum-capacity formula follows.
Require secrecy from a complementary output	An adversary and privacy criterion are added.	No private or secret-key formula follows.
Restrict the decoder to separate measurements	The allowed receiver class is smaller.	A new optimality proof would be needed.
Replace repeated GAD uses by an arbitrary partner	The quantifier over channels is enlarged.	Same-channel additivity is insufficient.
Use ordinary or unital damping only	A parameter subfamily is selected.	The full thermal range is not proved by specialization.
Replace independent uses by correlated bath noise	The channel need not be $A^{\otimes n}$ .	The memoryless coding model no longer applies unchanged.
Infer the coherence multiplier from population records	Only part of the channel has been measured.	The GAD structural promise or additional tomography is needed.
Use an area law, one-round game value, or oracle budget	The mathematical object and resource model change.	No channel-capacity identification is supplied.
Replace an asymptotic sequence by a finite confidence interval	The quantifier over block length is removed.	Coverage is not an asymptotic coding theorem.

General Holevo additivity is false; no statement in this manuscript concerns every quantum channel Hastings. The claims about external Families 265, 272, 273, 277, and 283 remain catalogue comparisons, not theorem premises Dossier276.