absorb.md

Massimo

Chronological feed of everything captured from Massimo.

paper / rainmaker1973 / Jun 13

Swap Regret Minimization Enables Efficient Coordination in Conflicted Multi-Agent Systems

Φ-Actor-Critic (Φ-AC) addresses the fundamental problem in general-sum games where Nash equilibria often fail to maximize collective welfare by leveraging swap regret minimization to steer learning toward Pareto-efficient correlated equilibria. The framework employs a centralized attention critic to tractably estimate counterfactual regrets in deep MARL, combined with Lagrangian-based equilibrium selection that balances social welfare optimization with stability constraints. Experimental validation across matrix games, MPE, and Harvest scenarios demonstrates the approach maintains both high collective returns and competitive fairness in mixed-motive settings.

multi-agent-reinforcement-learninggame-theorycorrelated-equilibriaactor-critic-methodscoordinationwelfare-optimizationregret-minimization
“Standard deep MARL methods converge to suboptimal Nash equilibria due to inherent limitations of value-decomposition monotonicity constraints and policy-gradient convergence patterns”
paper / rainmaker1973 / Jun 13

Superconductivity in Two-Orbital Iron Pnictides Requires Mott Insulating State, Absent in Reduced-Orbital Models

Variational Monte Carlo simulations of a two-orbital Hubbard-Kanamori model for iron-based superconductors reveal that s± superconductivity emerges only when the system exhibits Mott insulating behavior at half-filling, with no accompanying orbital selectivity. This contrasts sharply with three-orbital models that show extended superconducting regions paired with orbital selectivity effects, suggesting that orbital complexity—beyond minimal two-orbital treatments—is essential for capturing the physics of iron pnictides.

superconductivityhubbard-modeliron-pnictidesvariational-monte-carlostrongly-correlated-electronsorbital-selectivity
“A superconductive region with s± symmetry appears only when the half-filled system is in a Mott state”
paper / rainmaker1973 / Jun 10

Slitless Spectroscopy Algorithm Reconstructs 3D Flux Distributions for IFU-like Capabilities

This work introduces a novel non-parametric, data-driven algorithm for reconstructing 3D flux distributions from slitless spectroscopy by utilizing multiple dispersion angles. This method provides low-resolution IFU capabilities, effectively addressing source confusion without prior assumptions about object properties. The algorithm has been validated using simulated Roman Space Telescope data, demonstrating its ability to accurately subtract host galaxy light from transients and reconstruct detailed spectral features of galaxies, opening new avenues for large-scale, spatially resolved studies in astrophysics.

astronomydata-analysisspectroscopycosmologyinstrumentationsupernovae
“A non-parametric, data-driven algorithm can reconstruct 3D flux distributions from slitless spectroscopy, offering low-resolution IFU capabilities.”
paper / rainmaker1973 / Jun 10

Bounding Automorphism Groups of Algebraic Curves by $p$-Rank in Positive Characteristic

This paper addresses the challenge of bounding the automorphism group size of algebraic curves in positive characteristic $p$. It proves that unlike genera, no general bound exists based solely on the $p$-rank. However, a specific bound of the form $|\Gamma|<900 \left(\frac{p}{p-1}\right)^4 \gamma(\cX)^4$ is established under the condition that for a point $P$, the quotient curve $\cX/S_P$ is rational, where $S_P$ is the Sylow $p$-subgroup of the stabilizer of $P$ within $\Gamma$. This bound holds when $\gamma(\cX) \ge 2$ and is sharp up to the constant, with a caveat for subgroups of index at most 2 fixing $P$.

algebraic-geometryautomorphism-groupsp-rankalgebraic-curvesmathematics
“There is no universal function that bounds the automorphism group size of an algebraic curve solely based on its $p$-rank, unlike for genera.”
paper / rainmaker1973 / Jun 10

In-Situ Structural Analysis Reveals Dual-Stage Crystallization in Amorphous Tantala Thin Films

This study investigated the in-situ structural evolution of amorphous ion-beam sputtered tantala thin films during annealing using synchrotron radiation scattering. Key findings reveal a two-stage crystallization process: initial rapid formation of a cation backbone followed by progressive oxygen environment rearrangement, which enhances overall crystallinity. The methodologies employed, including XRD, Rietveld, and PDF analyses, provide detailed insights into these amorphous-to-crystalline transitions.

thin-filmsgravitational-wave-detectorsmaterials-sciencex-ray-diffractioncrystallizationoptics
“Ion-beam sputtered amorphous tantala films undergo structural rearrangements during annealing, featuring an initial rapid establishment of a cationic backbone structure.”