Model assessment ranking
AIRR asks the strongest suitable frontier models available for each assessment round to attack a paper, not merely summarize it. They search for counterexamples, hidden assumptions, proof gaps, unsupported novelty and reproducibility failures on the exact hashed version. AIRR promises no fixed provider, model or reasoning tier; founder-authored cases require two distinct identified models; every published score names the model, artifact, version and date, and unresolved material objections block admission.
A model may review a paper it helped develop, and its score counts. New rounds use separate conversations without development history, saved memory or other reviewers’ reports. Prior participation stays disclosed; legacy context that was not verified stays unknown.
Passing this unusually hard filter is meaningful positive evidence that a paper deserves serious attention. It is not infallibility: models can share blind spots, and a score cannot replace domain-expert review, formal proof or later correction.
- 01ARR-2026-58XZ9SCZA39DSRF0 · v1
Constant-gap entanglement distillation and a spectral criterion for noise-universal communication
7.00Exceptional · n=1 · 7.00–7.00Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 02ARR-2026-1DGGW5KX2M9R4VJT · v1
A universal capacity bound for amplified determinants and polynomial witnesses in the Hermitian case
6.00Excellent · n=1 · 6.00–6.00Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 03ARR-2026-418PW0QXGN812RDC · v1
Sharpness of the universal capacity bound for amplified determinants
5.80Excellent · n=1 · 5.80–5.80Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 04ARR-2026-4E4SJ328XF98BSCJ · v1
A dimension-independent capacity bound for amplified determinants of Hermitian matrix tuples
5.10Very good · n=1 · 5.10–5.10Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 05ARR-2026-68GTBBQT1T83QATJ · v1
A four-qubit criterion for entropy minimax duality: Minimality, sharp Renyi threshold and a W-state counterexample
5.00Very good · n=1 · 5.00–5.00Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 06ARR-2026-2GTKNVK5QM9H8BFR · v1
Universal optimal tests of unknown unitary dynamics: determinant exponents and tensor stability
4.93Very good · n=2 · 4.70–5.60Limited evidence · 1/2 benchmark matchesworking_paper · acceptance recorded separately - 07ARR-2026-68BH5JDJCQ8XW8GT · v1
One Fold, Unique Contact and the Complete Two-Piece Rate–Distortion Function of Rank-One Complex-Projective Born Prediction
4.85Very good · n=2 · 4.30–5.40Limited evidence · 0/2 benchmark matchesaccepted · acceptance recorded separately - 08ARR-2026-5QQF95VHTC9GABH8 · v1
Sharp Onset and Unbounded Growth of Norm-Optimal Self-Commutator Rank
4.71Very good · n=3 · 4.00–7.00Limited evidence · 1/3 benchmark matchesaccepted · acceptance recorded separately - 09ARR-2026-438PTRJA439VSRTW · v1
Exponential witness size and undecidable saturation for amplified determinants
4.70Very good · n=1 · 4.70–4.70Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 10ARR-2026-0VEP5BTYB99VNSC0 · v1
Six-spin Lorentzian obstructions for clique blowups of independence-stable graphs
4.70Very good · n=1 · 4.70–4.70Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 11ARR-2026-4C11Z7N0BX8VP8P2 · v1
Uniform tropical bounds for amplified determinants: finite phase witnesses and multiscale quantum noise
4.70Very good · n=1 · 4.70–4.70Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 12ARR-2026-7H9FAPTBZA897AMJ · v2
Complete Rate–Distortion Phase Diagram of Haar Oriented Two-Planes: One-Change Hypergeometric Coefficients, Unique Coexistence, No Reentrance, and Sharp Thermodynamic Contact Asymptotics
4.63Very good · n=3 · 4.20–4.80Limited evidence · 1/3 benchmark matchesaccepted · acceptance recorded separately - 13ARR-2026-5MAVBFTEF48DX8RT · v1
A complete class of optimal quantum tests for unknown unitary dynamics
4.60Very good · n=1 · 4.60–4.60Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 14ARR-2026-7993ZVQWNC9E38P2 · v1
Rank-two obstructions to Lorentzian graph partition functions
4.60Very good · n=1 · 4.60–4.60Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 15ARR-2026-3WVWP20DY09HFASA · v1
Rank plus Weyl sparsity is at least p + 1: an operator form of the Chebotarev-Tao uncertainty principle in prime dimension
4.58Very good · n=3 · 4.40–5.20Limited evidence · 1/3 benchmark matchesaccepted · acceptance recorded separately - 16ARR-2026-4Q1WFSE5DW9728QY · v1
The inverse self-commutator cost at inertia $(m,n)$: ordered-layering bounds, interlacing chains, and the Horn tiles that certify them
4.58Very good · n=2 · 4.50–4.80Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 17ARR-2026-1K33A7K90T87AREF · v1
The inverse self-commutator cost at inertia (m, 2): a dimension-free closed formula, block-shift optimizers and a Pieri certificate
4.55Very good · n=2 · 4.20–4.90Limited evidence · 0/2 benchmark matchesaccepted · acceptance recorded separately - 18ARR-2026-19WV17S3599GS8JR · v1
The Grassmannian Degree Law for Passive k-Plane Spectral Routing
4.55Very good · n=2 · 4.10–5.00Limited evidence · 0/2 benchmark matchesaccepted · acceptance recorded separately - 19ARR-2026-5KS70GV7KK9DYA69 · v1
Two-Query Chirality in Tetrahedral Quantum Echoes: Exact Parallel Readout, a Nine-Dimensional Query Defect, and a Certified Adaptive Advantage
4.54Very good · n=2 · 4.35–4.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 20ARR-2026-2Z86XAW5F88GQRZ2 · v1
The inverse self-commutator cost at inertia $(m,3)$: tiling certificates, 3×3-block shifts and a padding phase
4.53Very good · n=2 · 4.50–4.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 21ARR-2026-1VY960RN4T9E0RSC · v1
One Fold if and only if the Overlap Law is Right-Skewed: A Riccati Criterion for Tilted Beta Normalizers, with Consequences for Rank-One Projective Rate–Distortion
4.51Very good · n=2 · 4.00–6.00Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 22ARR-2026-1KTM07SEEV968BCB · v1
Sparse cyclic Weyl operators in prime-power dimension: an operator extension of the scalar uncertainty law through sparsity $2p-1$
4.51Very good · n=2 · 4.00–6.00Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 23ARR-2026-6XH6JAS5ZA934A6J · v1
Exact Classical Rate–Distortion for Phase-Lifted Generalized Coherent States: Cartan-Product Laplace Rigidity, Universal Radial Envelopes, and Slater-Determinant Transitions
4.51Very good · n=2 · 4.25–5.25Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 24ARR-2026-6FDEKPVJ0W8BHBMC · v1
A Finite-Dimensional Nonanalytic Spectral Transition and Exact High-Fidelity Rate–Distortion for Rank-r Born Prediction on Complex Grassmannians
4.50Very good · n=2 · 4.30–5.10Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 25ARR-2026-263B0753CQ9J2T34 · v1
Exact Branch Rigidity and Unique Crossing in Grassmann Matrix--Bingham Free Energies: Finite Certification, All-Field Gr_C(3,6) and Gr_C(4,8) Two-Block Order, and a Gr_C(2,5) Exchange Theorem
4.50Strong · n=2 · 4.20–4.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 26ARR-2026-0WESHW4YMM9FG8EK · v1
Three-term Weyl operators: central multiplicities, rank laws, and exact list decoding
4.49Strong · n=2 · 4.18–4.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 27ARR-2026-5B1F02F3EG9ZB9NN · v1
Exact Hive Certificates for the Balanced-Inertia Stability Constants (N = 3 to 12) in Every Dimension, the $(A_N,B)$ Cost Law, and the Twenty-Two-Form Six-Level Cost Formula
4.45Strong · n=3 · 4.00–6.00Limited evidence · 1/3 benchmark matchesaccepted · acceptance recorded separately - 28ARR-2026-1D2QYXPCVY9H7ANB · v1
Exact Rate–Distortion Theory for Complex-Projective Born Prediction: Finite-Dimensional Bingham Frontiers, Thermodynamic Coexistence, and Worst-State Capacity
4.41Strong · n=2 · 3.85–4.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 29ARR-2026-33GYTDAT5C8THB3M · v1
Phase calibration unlocks adaptive tests for renewed quantum noise
4.40Strong · n=1 · 4.40–4.40Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 30ARR-2026-61Y0FFA39M8KMBJ5 · v1
Complete Rank-Two Born-Prediction Rate–Distortion on Gr_C(2,4): All-Field Matrix–Bingham Rigidity and a Unique Coexistence Transition
4.39Strong · n=2 · 4.35–4.40Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 31ARR-2026-6F8XRSBM0J9Q2R2B · v1
All-Field Morse--Bott Stability at Critical Homogeneous Orbits: Half-Grassmann No-Spinodal Rigidity and Exact Jacobi Metastability
4.38Strong · n=2 · 3.75–4.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 32ARR-2026-7PEKNEDVH38HV8PX · v1
The arithmetic of sampling static quantum disorder: exact aliases, finite witnesses, and clock robustness
4.35Strong · n=2 · 4.20–4.40Limited evidence · 1/2 benchmark matchesworking_paper · acceptance recorded separately - 334.20Strong · n=1 · 4.20–4.20Limited evidence · 0/1 benchmark matchesaccepted · acceptance recorded separately
- 34ARR-2026-161TZ0CP488XTTBN · v1
Sparse Weyl operators: a uniform binary rank theorem and sharp erasure thresholds
4.20Strong · n=1 · 4.20–4.20Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 35ARR-2026-52B6MSS1W197W9T2 · v1
Exact Memory of Finite Spectral Routing Tables: The Direct-Sum Occupancy Law and Its Singular Strata
4.19Strong · n=2 · 4.00–4.75Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 36ARR-2026-6M3VGTXZ6W8JW9C9 · v1
Projective Memory and Resonant Bottlenecks in Passive Spectral Routing
4.16Strong · n=2 · 3.90–4.25Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 37ARR-2026-53CTRKDSP685PT51 · v5
The Schwarzian Bridge in a Single Sigmoid Neuron: Oracle Gram Resolution, Joint Spectral Lexicography, and Spherical Saturation Phases
4.15Strong · n=2 · 4.00–4.58Limited evidence · 1/2 benchmark matchescorrected · acceptance recorded separately - 38ARR-2026-6DJ302B1G38V4SHD · v1
Universal Semiclassical Coexistence in Classical Compression of Phase-Lifted Coherent States: Dimension-Normalized Contacts and a Matched High-Fidelity Boundary Layer
4.13Strong · n=2 · 3.65–4.30Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 39ARR-2026-6MMHJNP9E58DP9GB · v1
One Fold for Haar Oriented Two-Planes in Every Dimension n ≥ 6: A Phase-Space Proof through a Fano Factor, an Integrating Factor and the Threshold $L_q=q(q-1)/8$
4.10Strong · n=2 · 4.00–4.40Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 40ARR-2026-3H0ZKWJMH18MH9FX · v1
Strictly Scalable Exterior Decoders for Quantum Lists: Exact Full-Spark Widths and Fixed-Probe Weyl Bayes Curves
4.10Strong · n=2 · 4.10–4.10Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 41ARR-2026-7NPRNBW4488HG90K · v1
One-Spike Inverse Self-Commutators and Exact Three-versus-Four-Kick Curvature Synthesis
4.10Strong · n=1 · 4.10–4.10Limited evidence · 0/1 benchmark matchesaccepted · acceptance recorded separately - 42ARR-2026-6FWR12DH3F8QBA5C · v1
Sharp determinant bounds for amplified antisymmetric channels and a noise law for noncommutative rank
4.10Strong · n=2 · 3.80–4.20Limited evidence · 1/2 benchmark matchesworking_paper · acceptance recorded separately - 43ARR-2026-4VD8FW0NRZ8KGSNZ · v1
All-degree Born rigidity in dimensions four and three: an exact simplex optimum and a certified five-outcome frame
4.08Strong · n=2 · 4.00–4.30Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 44ARR-2026-4NC14QTZMT8708NN · v1
A complete one-state error–delay law for asymmetric three-node routing
4.07Strong · n=2 · 3.80–4.85Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 45ARR-2026-39Y0F83F4Z9JK90G · v2
The Diagonal Amos-Type Family at Real Order: a Machine-Checked Quantitative Crossing Classification
4.05Strong · n=2 · 4.00–4.20Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 46ARR-2026-54Q3HMFJ0Z8CZB4T · v1
Polynomial vertex reduction and optimal stability at balanced inertia (4,4)
4.02Strong · n=2 · 3.78–4.10Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 47ARR-2026-40S2JAQBBB8GJTN5 · v1
Constant-precision undecidability for amplified determinants with capacity one
4.00Strong · n=1 · 4.00–4.00Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 48ARR-2026-24M24KDPZK8HDBQ9 · v1
Sharp inertia ceilings and optimal stability for inverse self-commutators
4.00Strong · n=1 · 4.00–4.00Limited evidence · 1/1 benchmark matchesaccepted · acceptance recorded separately - 49ARR-2026-03PY6WTF258KGSRN · v1
Two-state exact routing: algebraic feasibility and sharp delay-two rigidity
3.96Strong · n=2 · 3.90–4.15Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 50ARR-2026-07EH3CZRD89YC8JN · v2
Lossless Calibration Is Stored Memory: A Topological McMillan-Degree and Wigner–Smith Law for Passive Quantum Networks
3.95Strong · n=2 · 3.80–4.40Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 51ARR-2026-7CCV86W3Y59VS8PN · v1
Matroidal Bayes Bounds for General Quantum Process Discrimination: Canonical Compression, Support Congestion, and Exact Qubit Phase Families
3.90Strong · n=2 · 3.80–4.20Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 52ARR-2026-6WX2JF38WE87GB2M · v1
Algebraic Query Support for Unitary Oracles: Exact Hilbert Laws, Harmonic Spectra, and General-Tester Bounds
3.89Strong · n=2 · 3.80–4.15Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 53ARR-2026-4FYHCBAQ0T8FHRBT · v1
Prescribed Tjurina Algebras and Complete Spectra in Osculating-Absorbing Gauss Fibres
3.87Strong · n=2 · 3.70–4.35Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 54ARR-2026-2MHNZRRJP49Y9SWP · v3
The Exact Rank Floor for Point-Span Tangent Absorption in Arbitrary Characteristic
3.87Strong · n=2 · 3.75–4.20Limited evidence · 1/2 benchmark matchescorrected · acceptance recorded separately - 55ARR-2026-15SJ1ANHDN8D88Z1 · v1
Bayesian Matroid-Union Bounds for Quantum List Discrimination: Support Congestion, Process Compression, and Exact Adaptive-Parallel Phases
3.84Strong · n=2 · 3.70–4.25Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 56ARR-2026-02AGP452839C09SH · v1
A finite geometry of low-rank three-list Weyl probes in ternary dimensions
3.83Strong · n=2 · 3.70–4.20Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 57ARR-2026-7X5XX0CBE19MXVSZ · v1
Finite-Sample-Valid Tests for Structured Matricial Hausdorff Moments: Joint Gaussian Grams, Strict Half-Time Separation, and Sharp Tangent-Cone Power
3.77Strong · n=2 · 3.60–4.25Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 58ARR-2026-19BXX42GM38K6VZ3 · v1
Characteristic-Free Bounds and Sharp Gauss-Fiber Examples for Point-Span Tangent Absorption
3.73Strong · n=2 · 3.60–4.10Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 59ARR-2026-77QM18J2KG9679B7 · v1
Finite-Sample Spectral-Gap Falsification: Exact Weighted Visibility Minimax, Hidden-Atom LAN, and Honest Dependent Tests
3.64Strong · n=2 · 3.40–4.35Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 60ARR-2026-6ZJY1SSSJA98WBWX · v1
Optimal stability and all equality cases at balanced inertia (5,5)
3.61Strong · n=2 · 3.60–3.62Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 61ARR-2026-7WNWKBJVQ28KZAKA · v1
Undecidable projection costs for quantum channels with finite tracial ancillas
3.60Strong · n=1 · 3.60–3.60Limited evidence · 0/1 benchmark matchesworking_paper · acceptance recorded separately - 62ARR-2026-0WAPCGQHNC82S8VJ · v1
Exact Multiplicity Floors for Dual Singularities from Absorbing Gauss Fibres
3.53Strong · n=2 · 3.25–4.35Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 63ARR-2026-31ED4MTKC18JPSDE · v2
Clustering and the Transfer-Operator Gap: A Machine-Checked Dense-Family Criterion
3.53Strong · n=2 · 3.30–4.20Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 64ARR-2026-5XEX8EX0629R997Y · v2
Euler-Reduced Tjurina Floors for Osculating-Absorbing Gauss Fibres
3.49Acceptable · n=2 · 3.40–3.75Limited evidence · 1/2 benchmark matchescorrected · acceptance recorded separately - 65ARR-2026-12W84G9BVC8BQAEQ · v1
Exact Rank Transitions through p=32 and a Half-Integral Optimum at p=53
3.48Acceptable · n=2 · 3.15–3.60Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 66ARR-2026-66Q8M61AA196T8BC · v2
Exact Floors and Proper-Span Extremizers for Higher Osculating Absorption
3.42Acceptable · n=2 · 3.25–3.90Limited evidence · 1/2 benchmark matchescorrected · acceptance recorded separately - 67ARR-2026-6TQRG7PY6C8JZR2D · v1
A sharp circle-root residual barrier and its collapse with one interior zero
3.40Acceptable · n=2 · 3.40–3.40Limited evidence · 0/2 benchmark matchesworking_paper · acceptance recorded separately - 68ARR-2026-0S8D4XJMZS9PDV1Z · v1
One calibration query activates a finite adaptive advantage in frozen-versus-renewed dephasing tests
3.40Acceptable · n=1 · 3.40–3.40Limited evidence · 1/1 benchmark matchesworking_paper · acceptance recorded separately - 69ARR-2026-2MZWECWVEN97ARVQ · v1
Fat Gauss Fibres and Tjurina–Milnor Defects Forced by Osculating Absorption
3.39Acceptable · n=2 · 3.30–3.65Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 70ARR-2026-1DNF7Z9F579X8SQA · v1
Exact Two-Dimensional SU(2) Yang-Mills in Lean: Weyl Integration, Heat-Kernel Convolution, Migdal Invariance, and the Exact Simple-Loop Area Law
3.35Acceptable · n=2 · 3.20–3.40Limited evidence · 1/2 benchmark matchesarchived · acceptance recorded separately - 71ARR-2026-3AJ7W5BS3N9VR8KM · v2
A Machine-Checked Proof of an Amos-Type Bound for Modified Bessel Ratios at Real Order
3.26Acceptable · n=2 · 3.00–4.00Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 72ARR-2026-2XFZW9FFEA8SXTXF · v1
Finite convex interpolation for polynomial Lyapunov existence
3.20Acceptable · n=2 · 3.00–3.40Limited evidence · 0/2 benchmark matchesworking_paper · acceptance recorded separately - 73ARR-2026-5286KSE6RS984S9P · v1
Interior-zero complexity of polynomial differentiation residuals
3.15Acceptable · n=2 · 3.00–3.30Limited evidence · 0/2 benchmark matchesworking_paper · acceptance recorded separately - 74ARR-2026-7W8BK4FK779W5ADJ · v1
A sharp polynomial Lyapunov threshold for a stable cubic family
3.10Acceptable · n=2 · 3.00–3.20Limited evidence · 0/2 benchmark matchesworking_paper · acceptance recorded separately - 75ARR-2026-5YJX444B6M95ZTG4 · v1
Classical capacities and complementary simulation for qubit Davies channels
3.10Acceptable · n=2 · 3.00–3.20Limited evidence · 0/2 benchmark matchesworking_paper · acceptance recorded separately - 76ARR-2026-7D2BBEC8MJ8BM80S · v1
Cellulation-Independent Boundary Gauge Averaging and Sharp Class-Sector Gaps in Two-Dimensional Yang--Mills
3.02Acceptable · n=2 · 2.90–3.35Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 77ARR-2026-3RH54K4NC58MRSX3 · v2
An exact relaxation optimum in a one-parameter family of four quantum measurement bases
2.90Acceptable · n=2 · 2.60–3.00Limited evidence · 1/2 benchmark matchesworking_paper · acceptance recorded separately - 78ARR-2026-3Q0VH7PMNV8C9VQE · v1
The Volume-Uniform Poincaré Walls: Machine-Checked Obstructions for Flat and Fluctuation-Sector Block-Poincaré Routes to Combes—Thomas Coercivity in Lattice Yang—Mills
2.75Acceptable · n=2 · 2.60–3.20Limited evidence · 1/2 benchmark matchesarchived · acceptance recorded separately - 79ARR-2026-7XT7AB8WJP9QPTGT · v1
Certified Localized Weil Positivity Through Support 0.72: Multiband Schur Complements and a Complete Stieltjes Hierarchy
2.35Substantial revision needed · n=2 · 2.20–2.40Limited evidence · 1/2 benchmark matchesaccepted · acceptance recorded separately - 80ARR-2026-6XR934XYNJ9V1B0P · v1
THE MASTER MAP: An Audit-First Navigation Guide to the Conditional Construction of 4D SU(N) Yang-Mills with Mass Gap - the Audited-Ledger
2.17Substantial revision needed · n=2 · 1.50–2.40Limited evidence · 1/2 benchmark matchesarchived · acceptance recorded separately - 81ARR-2026-1P88SB34HC8V492J · v1
Conditional Continuum Limit of 4d SU(Nc) Yang-Mills Theory via Two-Layer Architecture, RG-Cauchy Uniqueness, and Step-Scaling Confinement
1.95Substantial revision needed · n=2 · 1.80–2.00Limited evidence · 1/2 benchmark matchesarchived · acceptance recorded separately
See every paper and its assessment status
Five is very good, not a failing grade.
The 0.00–10.00 Millennium scale is not a school percentage or a probability of correctness. Three stars is the acceptable publication floor. Ten is the top comparison anchor and cannot be established by a model alone.
| Stars | Display | Meaning | Anchor |
|---|---|---|---|
| 1 | Critical concerns | ||
| 2 | Substantial revision needed | ||
| 3 | Acceptable | Publication floor | |
| 4 | Strong | ||
| 5 | Very good | Very good is deliberately above the publication floor | |
| 6 | Excellent | ||
| 7 | Exceptional | ||
| 8 | Potentially field-shaping | ||
| 9 | Potentially historic | ||
| 10 | Millennium-resolution benchmark | Unconditional recognized Millennium Problem solution after extraordinary verification |
How the score is calculated
Provisional method, AIRR-RATING-1.0. For a matching mathematical benchmark configuration, weight = 1 + 2 × max(0, accuracy − standard error). The score is the weighted mean of the original 0–10 scores. Weights stay between 1 and 3. Unknown model or effort receives weight 1 without an invented benchmark result. The raw report scores never change.
The pinned Epoch AI Tier 4 v2 snapshot, 12 September 2026 supplies mathematical-capability evidence; it does not measure reviewer honesty, novelty judgment or the probability that a paper is correct. The subtraction of one standard error and the weight cap are AIRR policy choices, not empirically calibrated reviewer reliability.
Evidence is described as Limited or Broader model support, with reasons on each paper. Models alone do not establish high confidence. Agreement, provider diversity, documented fresh blinded contexts and exact benchmark matches are recorded separately. Scores never overrule an unresolved material objection or sign an acceptance.
Criterion profile
Each report also supplies one to five stars, with a written basis, for correctness confidence, rigor, novelty, significance and reproducibility. These diagnostic ratings are shown separately and are not silently averaged into the headline score.
Prior manuscript participation does not exclude a report or reduce its numerical weight. AIRR includes one latest report per evidenced model, preserves every earlier report, and shows count, range, participation and review-context evidence. A new blinded thread is distinct from statistical independence. Missing reports receive no stars.
Download the versioned machine-readable assessment registry · JSON Schema
Assessment coverage: every current paper
Existing editorial status and assessment coverage are separate. A legacy acceptance does not invent a modern score; a private candidate report cannot be assigned to the public predecessor.