Skip to content

Category

generative ai

2,176 papers

#generative ai Open access Oct 2026

FDCL Part XI: Dynamical SU(2) Gauge Fields on FDCL Graphs — Hierarchical Curvature Completion and Local Spectral Bounds

We specify a dynamical SU(2) gauge model on finite FDCL voxel graphs, with Gauss invariance at every vertex. Unit plaquettes miss a central flux whose shortest detecting loop has length 2n + 4. An explicit hierarchical catalogue completes nonabelian flatness with the minimum number of added relations and the minimum po...

Bin Seol · 1 citation
#generative ai Open access Oct 2026

FDCL Part XII: Haar-Preserving Assembly in FDCL SU(2) Gauge Models — State Preparation, Shared-Link Responses, and Delayed Transfer

We study which information block assembly must retain to preserve responses in finite-graph SU(2) gauge theory. Scalar, pairwise and oriented triple invariants classify simultaneous conjugation orbits; we give exact consistency conditions including singular Gram ranks. Vertex multiplicities and physical path lengths de...

Bin Seol · 1 citation
#generative ai Open access Oct 2026

FDCL Part VII: Arithmetic Readouts and Prime-Set Structure in FDCL Models — Exact Obstructions and Finite-Field Reconstructions

We investigate what arithmetic information is retained by readouts of FDCL graphs, address automata and explicitly declared finite-field reconstruction models. A fixed additive length group cannot supply arbitrarily many independent prime logarithms, but a one-dimensional determinant sequence already has infinite multi...

Bin Seol · 3 citations
#generative ai Open access Oct 2026

FDCL Part VI: Zeta Functions and Trace Growth on FDCL Graphs — Exact Determinants and Infinite-Volume Bounds

We study metric spectral zeta functions, finite orbit determinants and nonbacktracking trace growth for explicitly distinguished FDCL graph models. For the unweighted Factory graph, edge-disjoint recursive copies make every anchor-normalized closed-walk count monotone and bounded. This proves existence of all fixed-len...

Bin Seol · 3 citations
#generative ai Open access Oct 2026

FDCL Part IV: Recursive Dynamics and Observation on FDCL Graphs — Finite-State Laws and Four-Mode Reduction

We study recursive generation and finite observation of FDCL graphs. Exact incidence and birth systems distinguish counting quotients from letterwise observation. Geometric restriction and lifting give shift equivalences between origin-pattern automata at different radii. The seed cyclic module has four persistent mode...

Bin Seol · 5 citations
#generative ai Open access Oct 2026

FDCL Part V: Gauge Hamiltonians and Scattering on FDCL Graphs — Exact Reductions and Observable Quotients

We develop exact operator reductions for several graph models associated with FDCL, keeping their underlying graphs, mass metrics and boundary conventions explicit. A harmonic extension yields a two-sided spectral-gap bound with its induced mass; a coordinate trial function rules out a proposed level-independent static...

Bin Seol · 6 citations
#generative ai Open access Oct 2026

FDCL Part VIII: Hierarchical Networks and Feedback Control in FDCL Models — Exact Routing, Stability and Observation Conditions

We develop explicit network, routing and feedback models motivated by FDCL geometry. Core–branch reductions preserve optimal congestion, while joint routing and capacity allocation on a general graph reduce to weighted shortest paths. A weighted graph field has an exact diffusion threshold including coupling to null mo...

Bin Seol · 6 citations
#generative ai Open access Oct 2026

FDCL Part II: Spectral Reduction and Renormalization on FDCL Graphs

We study spectral reduction on specified graphs associated with the six-digit FDCL construction. The coordinate-merged skeleton has an exact template spectrum and explicit all-level gap upper bounds for counting and degree-weighted mass. For the auxiliary eight-edge metric carrier, a pole-free matching equation gives t...

Bin Seol · 7 citations
#generative ai Open access Oct 2026

FDCL Part III: Resistance and Diffusion on FDCL Graphs — Exact Flows and Harmonic Reduction

We study resistance and diffusion on graph models associated with the six-digit FDCL construction. For the voxel graph, an explicit optimizing current gives the exact resistance between shorted horizontal faces. An orthogonal load decomposition and corrected planar currents prove that the prescribed uniform-face resist...

Bin Seol · 7 citations
#generative ai Open access Oct 2026

FDCL Part IX: Proof Methods and Exact Certification for FDCL Models — From Finite Evidence to Uniform Theorems

We develop explicit proof and certification interfaces for the FDCL program using classical geometric, linear-algebraic and finite-state methods. Reducible pressure bounds and sharp multiplicity examples separate symbolic growth from geometric dimension. Mass-orthogonal constraint repair, residual minimization and a no...

Bin Seol · 7 citations

From tech blogs

See all →
Microsoft Research Blog Oct 7, 2026

Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses

Training AI agents with reinforcement learning can be challenging because their tools, context, and decision-making are managed by complex frameworks. Agent Lightning connects existing agents to RL training, making it easier to improve them without rebuilding them. The post Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses appeared first on Microsoft Research.

Microsoft Research Blog Oct 6, 2026

What AI gets wrong and what failure teaches us

Jennifer Neville did not want to go into computer science—but that’s exactly where she landed. Neville discusses the starts and stops that led to her professional sweet spot and her work identifying “surprising failures” making it hard for AI to handle complexity.  The post What AI gets wrong and what failure teaches us appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.