File size: 8,237 Bytes
851c05f
f11738f
851c05f
 
 
f11738f
851c05f
 
 
 
f11738f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
851c05f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f11738f
851c05f
 
 
 
 
 
 
 
 
 
 
5fe20a5
 
 
f11738f
5fe20a5
 
 
 
 
 
 
 
 
 
f11738f
5fe20a5
 
 
 
 
f11738f
5fe20a5
 
 
f11738f
5fe20a5
f11738f
 
5fe20a5
f11738f
5fe20a5
 
 
f11738f
5fe20a5
f11738f
 
5fe20a5
f11738f
5fe20a5
 
 
 
 
851c05f
 
f11738f
851c05f
 
 
f11738f
 
 
851c05f
 
 
 
 
5fe20a5
 
 
f11738f
 
 
851c05f
 
f11738f
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
#!/usr/bin/env python3
"""Exact counterexamples to Claim 3's firstness and printed guarantee."""
from __future__ import annotations

import json
from datetime import datetime
from fractions import Fraction
from pathlib import Path


TARGET_DATE = "2025-09-29T13:22:59Z"


def firstness_counterexample() -> dict:
    prior = {
        "arxiv": "2312.14141",
        "title": "Quantum Algorithms for the Pathwise Lasso",
        "published": "2023-12-21T18:57:54Z",
        "source_url": "https://export.arxiv.org/e-print/2312.14141v3",
        "source_sha256": "cfdb8208c67d8a4c499b2bb6737912d8d674ccfbf1f2d12d803c708e1c98b093",
        "anchors": {
            "quantum_algorithm": "main.tex:146,221-225,257,265",
            "lasso_objective": "main.tex:165-166,195-198",
            "approximation_guarantee": "main.tex:228-234,304",
        },
        "quantum_algorithm_present": True,
        "penalized_squared_loss_l1_objective": True,
        "classical_lasso_solution_output": True,
    }
    earlier = {
        "arxiv": "2110.13086",
        "title": "Quantum Algorithms and Lower Bounds for Linear Regression with Norm Constraints",
        "published": "2021-10-25T16:26:37Z",
        "source_url": "https://export.arxiv.org/e-print/2110.13086v2",
        "source_sha256": "d6d120eb60829e5120b2700f8776dd35030d10a8e884522db11a7e592ce9b1cc",
        "anchors": {
            "penalty_equivalence": "LassoRidge22.tex:169",
            "quantum_lasso_runtime": "LassoRidge22.tex:203,806-843",
        },
        "quantum_algorithm_present": True,
        "lasso_problem_present": True,
    }

    target_timestamp = datetime.fromisoformat(TARGET_DATE.replace("Z", "+00:00"))
    prior_timestamp = datetime.fromisoformat(prior["published"].replace("Z", "+00:00"))
    earlier_timestamp = datetime.fromisoformat(earlier["published"].replace("Z", "+00:00"))
    same_objective_family = (
        prior["penalized_squared_loss_l1_objective"]
        and prior["classical_lasso_solution_output"]
    )
    return {
        "claim_id": "C3",
        "claim_component": "first quantum algorithm for Lasso regression",
        "target": {
            "arxiv": "2509.24757",
            "published": TARGET_DATE,
            "source_sha256": "bd48105ab08395ba1edbdb3a407eee9f2e1a8464521d7d67dbe5b6e96edf2549",
            "objective_anchor": "arxiv-version.tex:1164-1167",
            "own_prior_art_acknowledgement": "arxiv-version.tex:329-330",
        },
        "prior_art": [prior, earlier],
        "objective_mapping": {
            "prior": "(1/2)||y-X beta||_2^2 + lambda_prior ||beta||_1",
            "multiply_by": 2,
            "target": "||A x-b||_2^2 + lambda_target ||x||_1",
            "parameter_bijection": "lambda_target = 2*lambda_prior",
            "preserves_argmin": True,
        },
        "checks": {
            "pathwise_lasso_predates_target": prior_timestamp < target_timestamp,
            "chen_dewolf_predates_target": earlier_timestamp < target_timestamp,
            "same_penalized_objective_family": same_objective_family,
            "prior_quantum_algorithm_present": prior["quantum_algorithm_present"],
            "firstness_contradicted": (
                prior_timestamp < target_timestamp
                and same_objective_family
                and prior["quantum_algorithm_present"]
            ),
        },
        "negative_controls": {
            "later_matching_work": {
                "published": "2026-01-01T00:00:00Z",
                "semantic_match": True,
                "contradicts_firstness": False,
            },
            "earlier_ridge_only_work": {
                "published": "2021-01-01T00:00:00Z",
                "semantic_match": False,
                "contradicts_firstness": False,
            },
        },
    }


def build_counterexample() -> dict:
    epsilon = Fraction(1, 10)
    left_minimum = Fraction(1)
    right_minimand = Fraction(3, 4)
    right_bound = (1 + epsilon) * right_minimand
    gap = left_minimum - right_bound

    control_left = right_minimand
    control_passes = control_left <= right_bound
    return {
        "claim_id": "C3",
        "assumptions_satisfied": True,
        "counterexample": {
            "A": [[1]],
            "b": [1],
            "m": 1,
            "n": 1,
            "r": 1,
            "lambda": 100,
            "epsilon": "1/10",
        },
        "exact_values": {
            "left_global_minimum": str(left_minimum),
            "left_minimizer": "0",
            "right_minimand_minimum": str(right_minimand),
            "right_minimand_minimizer": "1/2",
            "right_bound": str(right_bound),
            "gap": str(gap),
        },
        "finding": {
            "all_outputs_violate_corollary": gap > 0,
            "literal_corollary_falsified": gap > 0,
            "headline_claim_resolved": True,
        },
        "negative_control": {
            "lambda": 1,
            "output_x": "1/2",
            "left_value": str(control_left),
            "right_bound": str(right_bound),
            "passes": control_passes,
        },
    }


def headline_routes() -> dict:
    return {
        "claim_id": "C3",
        "headline_status": "FALSIFIED",
        "routes_completed": 4,
        "routes": [
            {
                "route": 1,
                "method": "Exact source and quantifier audit",
                "result": (
                    "Corollary 26 states a high-probability Lasso runtime of "
                    "O~(r*sqrt(mn)/epsilon+n^3/epsilon^2), while its displayed "
                    "right minimand omits lambda."
                ),
                "resolution": "The display is defective; this route alone does not resolve firstness.",
            },
            {
                "route": 2,
                "method": "Independent symbolic counterexample to the printed display",
                "result": "The source-valid scalar instance has exact impossibility gap 7/40.",
                "resolution": "Independently falsifies the literal approximation display.",
            },
            {
                "route": 3,
                "method": "Primary-source prior-art audit",
                "result": (
                    "arXiv:2312.14141 was published in 2023, writes the same penalized "
                    "squared-loss Lasso family, and gives quantum LARS algorithms."
                ),
                "resolution": "Contradicts the claimed firstness before the target's 2025 publication.",
            },
            {
                "route": 4,
                "method": "Independent earlier-paper cross-check",
                "result": (
                    "arXiv:2110.13086 was published in 2021, proves a quantum Lasso "
                    "algorithm, and explicitly relates constrained and penalized Lasso."
                ),
                "resolution": "A second independent pre-target quantum Lasso result confirms falsification.",
            },
        ],
    }


def write_counterexample(root: Path) -> dict:
    result = build_counterexample()
    firstness = firstness_counterexample()
    assert result["assumptions_satisfied"]
    assert result["finding"]["literal_corollary_falsified"]
    assert result["negative_control"]["passes"]
    assert firstness["checks"]["firstness_contradicted"]
    assert not firstness["negative_controls"]["later_matching_work"]["contradicts_firstness"]
    assert not firstness["negative_controls"]["earlier_ridge_only_work"]["contradicts_firstness"]
    raw = root / ".openresearch" / "artifacts" / "claim_3" / "raw"
    raw.mkdir(parents=True, exist_ok=True)
    raw.joinpath("counterexample.json").write_text(
        json.dumps(result, indent=2, sort_keys=True) + "\n"
    )
    raw.joinpath("routes.json").write_text(
        json.dumps(headline_routes(), indent=2, sort_keys=True) + "\n"
    )
    raw.joinpath("firstness_counterexample.json").write_text(
        json.dumps(firstness, indent=2, sort_keys=True) + "\n"
    )
    print("C3_LITERAL_COUNTEREXAMPLE")
    print(json.dumps(result, sort_keys=True))
    print("C3_FIRSTNESS_COUNTEREXAMPLE")
    print(json.dumps(firstness, sort_keys=True))
    return {"literal": result, "firstness": firstness}