Skip to content
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
45 commits
Select commit Hold shift + click to select a range
bcd5e3f
Add witness-budget experiments and modulator-only counterexamples
AwakeningOS Oct 1, 2026
3190e5d
Formalize the short-product budget counterexample for CI verification
AwakeningOS Oct 2, 2026
64c6650
Publish short-product candidate falsification and complete certificates
AwakeningOS Oct 2, 2026
fd44896
Respond to independent review: gcd closure, box witnesses and u=1 bounds
AwakeningOS Oct 3, 2026
6e513c7
Falsify fixed-small-shift escape with complete a ranges and Lean bounds
AwakeningOS Oct 4, 2026
a51cc44
Audit reply 27, correct minimum shift, and verify Lean bridge
AwakeningOS Oct 5, 2026
4d440f3
Review PR6 and refute nonresidue logarithmic budget candidate
AwakeningOS Oct 6, 2026
3c664c5
research: validate Jemmy witnesses and CRT/complementarity limits
AwakeningOS Oct 7, 2026
0881d06
research: test Type complementarity and ordinary-CF candidates
AwakeningOS Oct 8, 2026
fe2150e
research: independently review PR7 specification and fixed failure re…
AwakeningOS Oct 9, 2026
2fda2fe
Research: refute logarithmic Kneser-budget strengthening at p8615161
AwakeningOS Oct 10, 2026
139e960
Research review: Lean source checks, two fixed Kneser boxes and split…
AwakeningOS Oct 11, 2026
09a6f58
research: test fixed-p absolute character budget and certify finite f…
AwakeningOS Oct 12, 2026
b84bbb7
research: correct clipped residual identity and classify complex fini…
AwakeningOS Oct 13, 2026
4e514ac
Research: audit certificate slack and exact character-orbit traces
AwakeningOS Oct 14, 2026
ddf9066
Research: split square-coset failures and refute odd-only log budget
AwakeningOS Oct 15, 2026
aa0c3d1
Research: distinguish forced zero-slack families and outcome-driven a…
AwakeningOS Oct 16, 2026
4da566a
Research: refute log exact-family restrictions and test phase-discard…
AwakeningOS Oct 17, 2026
2eb1317
Clarify automatic repository CI versus unformalized research bound
AwakeningOS Oct 17, 2026
4977cf0
Document closure audit and finite-exponent bridge; verify reciprocity…
AwakeningOS Oct 18, 2026
41200f8
Audit moving divisor sets and CRT transfer; correct L8 CSV manifest
AwakeningOS Oct 19, 2026
98c676a
research: independently audit N8 minimum input and uploaded data cons…
AwakeningOS Oct 21, 2026
8313f47
research: audit exponent deficits and check conditional Type II bridg…
AwakeningOS Oct 22, 2026
34bf056
research: audit square shifts and formalize power-preservation limits
AwakeningOS Oct 23, 2026
c6a0617
research: refute sufficient-short-root-only escape with finite Lean c…
AwakeningOS Oct 25, 2026
e6acb78
research: seed typed witness and auxiliary-counterexample registry
AwakeningOS Oct 26, 2026
16ff5ac
Research: refute all-vertex one-edge rescue with a Lean-checked finit…
AwakeningOS Oct 27, 2026
d544381
Independently replay reported E1 counterexample at p=42698041; retire…
AwakeningOS Oct 29, 2026
9b05e55
Sharpen prime-modulus Type II obstruction deduction using Euler crite…
AwakeningOS Oct 29, 2026
4c1c3e4
Prove sigma3 Type II base-group criterion and diagnose fixed root data
AwakeningOS Oct 30, 2026
daa8731
Audit radical reduction scope and bounded-exponent lifting obstruction
AwakeningOS Oct 31, 2026
4662076
Document and verify a source-to-child projection counterexample
AwakeningOS Nov 4, 2026
fbe28d8
Audit multiplier changes and residual failure at the fixed child 73
AwakeningOS Nov 5, 2026
7aaf51d
Clarify component quantifiers and retain full-square closure in the p…
AwakeningOS Nov 6, 2026
c2345d4
Audit the 167 reported branches and test canonical integer-ratio witn…
AwakeningOS Nov 7, 2026
b4ce605
Audit rational Type I witnesses and a depth-two endpoint for 280009
AwakeningOS Nov 8, 2026
f3008bb
Refute canonical integer-ratio terminal coverage at p=30241
AwakeningOS Nov 9, 2026
76b79cd
Audit successful sources and the scope of canonical selector transfers
AwakeningOS Nov 10, 2026
e1058e1
Verify reply cohorts and distinguish canonical parent edges from chil…
AwakeningOS Nov 11, 2026
f1169a2
Verify fixed-point tables and canonical terminal restriction at p30241
AwakeningOS Nov 12, 2026
d4de52c
research: refute extremal-terminal rule on a full-square closed compo…
AwakeningOS Nov 13, 2026
a99b1f0
research: audit reply139 and separate connectivity from selector exis…
AwakeningOS Nov 14, 2026
8399e66
research: test prime-source closed cycles in response to post141
AwakeningOS Nov 15, 2026
7956611
research: certify reduced large-prime source obstruction from reply143
AwakeningOS Nov 16, 2026
20b7ac7
research: verify symbolic upper-square-source character obstruction
AwakeningOS Nov 17, 2026
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
4 changes: 4 additions & 0 deletions OutcastsMathLab.lean
Original file line number Diff line number Diff line change
@@ -1 +1,5 @@
import OutcastsMathLab.Problems.P000_VerifierSmoke.Proof
import OutcastsMathLab.Research.ShortProduct20261002
import OutcastsMathLab.Research.Reply20261003
import OutcastsMathLab.Research.JointShift20261004
import OutcastsMathLab.Research.Reply20261005
61 changes: 61 additions & 0 deletions OutcastsMathLab/Research/JointShift20261004.lean
Original file line number Diff line number Diff line change
@@ -0,0 +1,61 @@
import OutcastsMathLab.Research.Reply20261003

/- A completeness lemma used to falsify the proposed u=1-or-2 escape route.
It is NOT a pointwise existence theorem. The new finite negative rows are
independently audited in Python, not assumed or encoded as Lean axioms. -/
namespace OutcastsMathLab.Research.JointShift20261004

open OutcastsMathLab.Research.Reply20261003
set_option maxHeartbeats 4000000

/-- Every fixed-u actual witness has a finite a search bound, with no
primality, coprimality, or prime-factor-count assumption. -/
theorem fixed_shift_bound (p a u s : ℕ) (ha : 0<a) (hu : 0<u)
(hs : 0<s) (hw : DivisorWitness p a u s) :
4*u*a ≤ p+4*u*u+1 := by
obtain ⟨v,heq⟩ := hw.1
have hsv : s*v = a*p+u := by simpa using heq.symm
have hv : 0<v := by nlinarith
have hslo : 4*a*u ≤ s+1 := Nat.le_of_dvd (by omega) hw.2
have had : a ∣ s+1 := dvd_trans (show a ∣ 4*a*u from ⟨4*u,by ring⟩) hw.2
have hatotal : a ∣ a*p+(u+v) := by
have ht : a ∣ (s+1)*v := dvd_mul_of_dvd_left had v
convert ht using 1; nlinarith [hsv]
have hav : a ∣ u+v := (Nat.dvd_add_iff_right (dvd_mul_right a p)).mpr hatotal
have hvlo : a ≤ u+v := Nat.le_of_dvd (by omega) hav
by_cases hau : a ≤ u
· nlinarith
· have hsl : 4*a*u-1 ≤ s := by omega
have hvl : a-u ≤ v := by omega
have hprod := Nat.mul_le_mul hsl hvl
have hmu : 0<4*a*u := Nat.mul_pos (by omega : 0<4*a) hu
have hfour : 4*a*u-1+1=4*a*u := Nat.sub_add_cancel (by omega)
have hone : a-u+u=a := by omega
by_contra hn
have hp : p+4*u*u+2 ≤ 4*u*a := by omega
have ht := Nat.mul_le_mul_left a hp
nlinarith [hprod,hsv,ht]

theorem fixed_shift_search_complete (p a u s : ℕ) (ha : 0<a) (hu : 0<u)
(hs : 0<s) (hw : DivisorWitness p a u s) :
a ≤ (p+4*u*u+1)/(4*u) := by
exact (Nat.le_div_iff_mul_le (by omega : 0<4*u)).mpr
(by simpa [Nat.mul_comm] using fixed_shift_bound p a u s ha hu hs hw)

/-- Concrete rescue at u=4, outside product log-budget 17.
The all-depth u=2,3 negative results are not formalized by this theorem. -/
theorem rescue_66529 :
Nat.Prime 66529 ∧ (66529:ℕ)%840=169 ∧
DivisorWitness 66529 11 4 703 ∧ (703:ℕ)=19*37 ∧
11*66529+4=703*1041 ∧ 703+1=4*11*4*4 ∧
DivisorWitness 66529 1 5 39 ∧ (17:ℕ)<11*4 ∧
(2:ℕ)^16<66529 ∧ 66529≤(2:ℕ)^17 ∧
4*(16656:ℕ)*11709104*3047294316 =
66529*(11709104*3047294316+16656*3047294316+16656*11709104) := by
norm_num [DivisorWitness]

#print axioms fixed_shift_bound
#print axioms fixed_shift_search_complete
#print axioms rescue_66529

end OutcastsMathLab.Research.JointShift20261004
117 changes: 117 additions & 0 deletions OutcastsMathLab/Research/Reply20261003.lean
Original file line number Diff line number Diff line change
@@ -0,0 +1,117 @@
import OutcastsMathLab.Research.ShortProduct20261002

/- Response to Outcasts #22 and #24. These are validation and closure lemmas,
not a proof of pointwise Type II existence or of the conjecture. -/
namespace OutcastsMathLab.Research.Reply20261003

open OutcastsMathLab.Research.ShortProduct20261002

set_option maxHeartbeats 4000000

def DivisorWitness (p a u s : ℕ) : Prop :=
s ∣ a*p+u ∧ 4*a*u ∣ s+1

/-- Dahan v1 Lemma 3.10: cancel a common factor without changing s. -/
theorem divisor_witness_cancel_factor (p d a u s : ℕ)
(hw : DivisorWitness p (d*a) (d*u) s) :
DivisorWitness p a u s := by
have hdmod : d ∣ 4*(d*a)*(d*u) := ⟨4*d*a*u, by ring⟩
have hds : d ∣ s+1 := dvd_trans hdmod hw.2
have hcs : Nat.Coprime s (s+1) := by simp
have hcd : Nat.Coprime s d := hcs.of_dvd_right hds
have hlin : s ∣ d*(a*p+u) := by
convert hw.1 using 1; ring
have hsmallmod : 4*a*u ∣ 4*(d*a)*(d*u) := ⟨d*d, by ring⟩
exact ⟨hcd.dvd_of_dvd_mul_left hlin, dvd_trans hsmallmod hw.2⟩

theorem short_witness_cancel_factor (p d a u : ℕ)
(hw : ShortWitness p (d*a) (d*u)) : ShortWitness p a u := by
rcases hw with ⟨q,hq,hd,hm⟩ | ⟨q,r,hq,hr,hd,hm⟩
· have h := divisor_witness_cancel_factor p d a u q
⟨hd, Nat.dvd_of_mod_eq_zero hm⟩
exact Or.inl ⟨q,hq,h.1,Nat.mod_eq_zero_of_dvd h.2⟩
· have h := divisor_witness_cancel_factor p d a u (q*r)
⟨hd, Nat.dvd_of_mod_eq_zero hm⟩
exact Or.inr ⟨q,r,hq,hr,h.1,Nat.mod_eq_zero_of_dvd h.2⟩

/-- Normalization preserves the same prime or prime-product witness and does
not increase a product budget. -/
theorem short_witness_normalize (p a u B : ℕ) (ha : 0<a) (hu : 0<u)
(hb : a*u ≤ B) (hw : ShortWitness p a u) :
∃ a' u', 0<a' ∧ 0<u' ∧ Nat.Coprime a' u' ∧
a'*u' ≤ B ∧ ShortWitness p a' u' := by
let d := Nat.gcd a u
have haeq : d*(a/d) = a := Nat.mul_div_cancel' (Nat.gcd_dvd_left a u)
have hueq : d*(u/d) = u := Nat.mul_div_cancel' (Nat.gcd_dvd_right a u)
have hsmall : ShortWitness p (a/d) (u/d) := by
apply short_witness_cancel_factor p d (a/d) (u/d)
simpa only [haeq,hueq] using hw
refine ⟨a/d,u/d,Nat.div_gcd_pos_of_pos_left u ha,
Nat.div_gcd_pos_of_pos_right a hu,?_,?_,hsmall⟩
· exact Nat.gcd_div_gcd_div_gcd_of_pos_left ha
· exact le_trans (Nat.mul_le_mul (Nat.div_le_self a d) (Nat.div_le_self u d)) hb

/-- Closes the noncoprime gap noted in #22: ALL positive pairs, not just 63. -/
theorem no_short_witness_budget23_all (a u : ℕ) (ha : 0<a) (hu : 0<u)
(hb : a*u ≤ 23) : ¬ShortWitness 5843041 a u := by
intro hw
obtain ⟨a',u',ha',hu',hc,hb',hw'⟩ :=
short_witness_normalize 5843041 a u 23 ha hu hb hw
exact no_short_witness_budget23 a' u' ha' hu' hb' hc hw'

/-- The same prime has semiprime witnesses in a BOX of side 23, outside the
product budget. Neither the box hypothesis nor ESC is refuted here. -/
theorem box_short_witnesses :
(578519:ℕ)=23*25153 ∧ ¬Nat.Prime 578519 ∧
DivisorWitness 5843041 10 9 578519 ∧
(116087:ℕ)=29*4003 ∧ ¬Nat.Prime 116087 ∧
DivisorWitness 5843041 3 14 116087 ∧
23 < (10:ℕ)*9 ∧ 23 < (3:ℕ)*14 := by
norm_num [DivisorWitness]

theorem box_short_witnesses_predicate :
ShortWitness 5843041 10 9 ∧ ShortWitness 5843041 3 14 := by
constructor
· exact Or.inr ⟨23,25153,by norm_num,by norm_num,by norm_num,by norm_num⟩
· exact Or.inr ⟨29,4003,by norm_num,by norm_num,by norm_num,by norm_num⟩

/-- #24's finite search bound, with no primality assumption on p.
It applies to ANY actual divisor witness, not merely a short product. -/
theorem modulator_only_bound (p a s : ℕ) (ha : 0<a)
(hs : 0<s) (hw : DivisorWitness p a 1 s) : 4*a ≤ p+5 := by
obtain ⟨v,heq⟩ := hw.1
have hsv : s*v = a*p+1 := by simpa using heq.symm
have hv : 0<v := by nlinarith
have hmod : 4*a ∣ s+1 := by simpa using hw.2
have hslo : 4*a ≤ s+1 := Nat.le_of_dvd (by omega) hmod
have had : a ∣ s+1 := dvd_trans (show a ∣ 4*a from ⟨4,by ring⟩) hmod
have hatotal : a ∣ a*p+(v+1) := by
have ht : a ∣ (s+1)*v := dvd_mul_of_dvd_left had v
convert ht using 1; nlinarith [hsv]
have hav : a ∣ v+1 := (Nat.dvd_add_iff_right (dvd_mul_right a p)).mpr hatotal
have hvlo : a ≤ v+1 := Nat.le_of_dvd (by omega) hav
have hsl : 4*a-1 ≤ s := by omega
have hvl : a-1 ≤ v := by omega
have hprod := Nat.mul_le_mul hsl hvl
have hfour : 4*a-1+1=4*a := by omega
have hone : a-1+1=a := by omega
by_contra hn
have hp : p+6 ≤ 4*a := by omega
have ht := Nat.mul_le_mul_left a hp
nlinarith [hprod,hsv,ht]

theorem modulator_only_search_complete (p a s : ℕ) (ha : 0<a)
(hs : 0<s) (hw : DivisorWitness p a 1 s) : a ≤ (p+5)/4 := by
exact (Nat.le_div_iff_mul_le (by norm_num : 0<(4:ℕ))).mpr
(by simpa [Nat.mul_comm] using modulator_only_bound p a s ha hs hw)

#print axioms divisor_witness_cancel_factor
#print axioms short_witness_cancel_factor
#print axioms short_witness_normalize
#print axioms no_short_witness_budget23_all
#print axioms box_short_witnesses
#print axioms box_short_witnesses_predicate
#print axioms modulator_only_bound
#print axioms modulator_only_search_complete

end OutcastsMathLab.Research.Reply20261003
53 changes: 53 additions & 0 deletions OutcastsMathLab/Research/Reply20261005.lean
Original file line number Diff line number Diff line change
@@ -0,0 +1,53 @@
import OutcastsMathLab.Research.JointShift20261004

/- Reply to Outcasts #27. The bounds are validation helpers, not new
existence results. No finite negative enumeration is encoded as an axiom. -/
namespace OutcastsMathLab.Research.Reply20261005

open OutcastsMathLab.Research.Reply20261003
open OutcastsMathLab.Research.JointShift20261004
set_option maxHeartbeats 4000000

/-- A divisor of the positive linear form cannot be zero. -/
theorem witness_s_pos (p a u s : ℕ) (hu : 0<u)
(hw : DivisorWitness p a u s) : 0<s := by
have hpos : 0<a*p+u := by omega
exact Nat.pos_of_dvd_of_pos hw.1 hpos

/-- The old fixed-shift bound without its redundant s-positive hypothesis. -/
theorem fixed_shift_bound_auto (p a u s : ℕ) (ha : 0<a) (hu : 0<u)
(hw : DivisorWitness p a u s) : 4*u*a ≤ p+4*u*u+1 :=
fixed_shift_bound p a u s ha hu (witness_s_pos p a u s hu hw) hw

theorem fixed_shift_search_complete_auto (p a u s : ℕ) (ha : 0<a) (hu : 0<u)
(hw : DivisorWitness p a u s) : a ≤ (p+4*u*u+1)/(4*u) :=
fixed_shift_search_complete p a u s ha hu (witness_s_pos p a u s hu hw) hw

/-- p has an actual u=1 witness. It is NOT a second u=1-or-2 counterexample.
Minimality of a=30 and C=12 are separately checked finite results. -/
theorem witnesses_1740481 :
Nat.Prime 1740481 ∧ (1740481:ℕ)%840=1 ∧
(2:ℕ)^20<1740481 ∧ 1740481≤(2:ℕ)^21 ∧
DivisorWitness 1740481 30 1 586679 ∧
30*1740481+1=586679*89 ∧ 586679+1=4*30*4889 ∧
(21:ℕ)<30 ∧
DivisorWitness 1740481 104 3 11231 ∧
(11231:ℕ)=11*1021 ∧ 104*1740481+3=11231*16117 ∧
11231+1=4*104*3*9 ∧
DivisorWitness 1740481 3 4 3311 ∧
3*1740481+4=3311*1577 ∧ 3311+1=4*3*4*69 ∧
(3:ℕ)*4≤21 ∧
4*(435121:ℕ)*255276348270*22719594996030 =
1740481*(255276348270*22719594996030+
435121*22719594996030+435121*255276348270) ∧
4*(435159:ℕ)*4887270648*26256047011272 =
1740481*(4887270648*26256047011272+
435159*26256047011272+435159*4887270648) := by
norm_num [DivisorWitness]

#print axioms witness_s_pos
#print axioms fixed_shift_bound_auto
#print axioms fixed_shift_search_complete_auto
#print axioms witnesses_1740481

end OutcastsMathLab.Research.Reply20261005
Loading
Loading