PAPERS / RESEARCH NOTES
Read world-model, mathematical and foundational research manuscripts, with evidence, limits and possible next steps.
Working manuscripts, formal verification records and research handoffs are presented with their individual status. Public availability does not imply peer review or external recognition.
Complete Hugging Face archive ↗31 items
WORLD-MODELS
Exact small environments separate parameter information, belief error and task value; reproducible classical Bayes diagnostics, not a new exploration principle or visual-model capability claim.
WORLD-MODELS
With the same forecaster and per-decision budget, a CEM objective change raises success from 11/50 to 33/50; a task-specific objective effect, not a new dynamics mechanism or generalization claim.
WORLD-MODELS
Studies calibration during actuator shifts and its restoration cost; improving an old adaptive reference can still harm frozen-model performance, and learned control superiority is not established.
WORLD-MODELS
Explains why the conditional-mean squared-error identity does not certify physical identification, calibrated posterior weights or control optimality.
WORLD-MODELS
Records real packed parameters and recoverable updates alongside negative action diagnostics, extremely few optimization steps and measured device limits; no trained general-model capability is claimed.
WORLD-MODELS
Runs and adapts a released tri-view predictor on limited hardware; local error reduction still trails static persistence, and the historical scored system is distinct from the adapter.
WORLD-MODELS
A separate 11,285-parameter prototype improves reused-set pixel error, while EEF removal gives similar performance; action contribution, unseen-task validity and artifact binding remain unresolved.
WORLD-MODELS
Reports both geometric/temporal strengths and task/motion weaknesses of a historical system; a descriptive score profile, not proof of a new method, physical-robot performance or general planning.
WORLD-MODELS
Separates cost projection from sparse prediction; paid calibration and later schedules fail promotion gates, and a single-seed signal fails replication. An internal candidate claim was withdrawn, not an external paper retracted.
WORLD-MODELS
Exact finite causal diagnostics and references; technical report.
WORLD-MODELS
Ten-seed bounded synthetic diagnostic; technical report.
WORLD-MODELS
Finite-history guarantees and counterexamples under explicit assumptions; research notes.
WORLD-MODELS
Action-path audit, parameter diagnostics, classical net-value repair and stopping criteria; research notes.
MATHEMATICS
A 5×4 construction, genus-13 reduction, moving genus-three curve and squareclass-growth note.
MATHEMATICS
A 5×4 construction, genus-13 reduction, moving genus-three curve and squareclass-growth note.
PHYSICS
Finite-product dimension bounds, moving-pole scalar dominance and a sharp-gap supplement, with prior constructions credited.
PHYSICS
Positive-matrix parametrization and purity thresholds under stated extension assumptions, using the manuscript normalization.
PHYSICS
A proposed global induction and rotation-singlet improvement, with a coefficient audit and its clarification.
PHYSICS
A proposed global induction and rotation-singlet improvement, with a coefficient audit and its clarification.
PHYSICS
A proposed global induction and rotation-singlet improvement, with a coefficient audit and its clarification.
PHYSICS
A low-particle gap manuscript and harmonic-moment stability supplement, separating prior qualitative results from proposed quantitative steps.
PHYSICS
A low-particle gap manuscript and harmonic-moment stability supplement, separating prior qualitative results from proposed quantitative steps.
MATHEMATICS
A first-question manuscript and all-n formalization, with a later Lean supplement for the asymptotic tripartite second question and conjunction.
MATHEMATICS
A 5×4 construction, genus-13 reduction, moving genus-three curve and squareclass-growth note.
PHYSICS
Finite-product dimension bounds, moving-pole scalar dominance and a sharp-gap supplement, with prior constructions credited.
PHYSICS
Finite-product dimension bounds, moving-pole scalar dominance and a sharp-gap supplement, with prior constructions credited.
MATHEMATICS
Catalogue-conditional deductions and partial degree-ten positive witnesses, not a complete enumeration of all graphs.
PHYSICS
Exact identities, reductions and counterexamples to overstrong relaxations.
ASTRONOMY
Sample-alignment and provenance diagnostics for pinned inputs.
CRYPTOGRAPHY
Specified incidence instances, assumptions and finite witnesses; no production-network security estimate.
MATHEMATICS
Three-point spectral-defect bounds, a second-moment saturation obstruction and finite interval probes of Weil kernels.
No matching manuscripts.
