Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.03243

A Kinetic Theory of the Gated Self-Evolving LLM Agent

Haipeng Wang
View original source ↗Read source document ↗Export metadata ↓

Abstract

We find traces of fluid dynamics in the self-evolution of an LLM agent, and give the kinetic theory that predicts them. Gated self-evolution is the loop in which an agent rewrites its own skills under a validation gate. Self-evolution research has treated the agent as the unit; we study instead the individual instances inside it. Here the agent is DSH-plugin-based: it runs in production on DeepSeek Harness (DSH), and its plugins satisfy four architectural properties (permutation symmetry, reversibility, acyclicity, typed contracts), which license treating these instances as identical hard spheres; the theory is accordingly scoped to DSH-class plugin populations. On this scope the paper builds three theory layers. The rigorous layer, independent of any analogy, comprises an any-time hitting-time certificate bounding the expected rounds to any prescribed improvement, a resolution law that prices held-out validation budgets, and a separation theorem: the daemon must stay outside the population, because merging evaluator with evaluated voids the certificate. The kinetic layer is a master equation over the plugin x version x task grid with four operators (collision, reaction, external field, gate), where collision is co-activation. Its moment hierarchy, the step that turns a gas into fluid equations, generates the falsifiable statistical signatures. Throughout, the fluid reading is a bounded analogy: momentum is not conserved, so no Navier-Stokes limit exists. The measured layer runs on a faithful minimal instance, a large library of four-parameter skill plugins retrieved one per episode with a co-activation probe, in a one-model, one-task-family WebShop environment; every element maps to the DSH loop by architectural role. Population fluctuation scaling is density-gated: invisible at sparse edit density, it emerges at the predicted rate under tripled density, as directional evidence.

Explore related subjects

cs.GR

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Haipeng Wang. 2026-10-02. A Kinetic Theory of the Gated Self-Evolving LLM Agent. https://arxiv.org/abs/2610.03243

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Fourier-Latent Diffusion for Constrained Generation of Triply Periodic Minimal Surfaces

We present a generative framework for the constrained design of $D_{2h}$-symmetric triply periodic minimal surfaces (TPMS) with low residual mean curvature. Existing TPMS design pipelines either explore a limited set of canonical analytical families or rely on computationally expensive numerical procedures, making it difficult to efficiently generate diverse candidates under geometric and/or mechanical requirements. Our approach learns a distribution of solver-generated TPMS in a compact Fourier space where periodicity and $D_{2h}$ symmetry are guaranteed by construction. This representation eliminates the need for a learned geometric decoder and enables the training of an effective diffusion model that generates low-mean-curvature candidates conditioned on sparse geometric constraints, selected homogenized elastic properties, or their combination. A subsequent coefficient-space refinement further reduces the residual mean curvature. Experiments show that our approach outperforms trigonometric, SDF-based, and standard Fourier-space baselines and enables controllable, high-quality generation under geometric and low-dimensional mechanical constraints. Overall, the proposed framework provides a compact and efficient design space for generating near-minimal periodic structures under user-specified requirements.

cs.GR↗
arXiv2026

ARS-Avatar: Animatable and Relightable Surfel Avatars with Learnable Ambient Occlusion

Creating animatable and relightable human avatars from multi-view images remains challenging, as pose-dependent deformation, materials, and light visibility are tightly coupled in images. In this paper, we present ARS-Avatar, a novel method using surfel representation for high-quality, animatable, and relightable human avatars from multi-view images captured under unknown illumination. We first extract deformation priors from the template mesh and leverage them as additional guidance beyond driving poses, facilitating faithful estimation of surfel attributes. To support relighting, we employ deferred shading to estimate BRDF materials. We further introduce a differentiable screen-space ambient occlusion formulation that enables gradient-based optimization of body-part specific occlusion radii through finite differences, providing an efficient approximation of light visibility that can be jointly optimized with the avatar. Extensive experiments show that ARS-Avatar achieves competitive or improved radiance reconstruction on multiple metrics and consistently outperforms the evaluated PBR relighting baselines, while enabling realistic animation and relighting under novel poses and illuminations.

cs.GR↗
arXiv2026

Computational Design of Free-Form Lampshades

Conventional lampshades are predominantly restricted to simple geometries, owing to the difficulty of controlling optical functions across complex shapes. We present a computational design method that enables 3D-printable lampshades with complex free-form shapes to achieve controlled illumination, such as uniform luminance. Given an arbitrary shape, our method determines the internal cavity geometry and the number and placement of lights in two stages: an initialization stage fixes the number of lights and establishes their initial positions and the initial cavity geometry, after which gradient-based optimization jointly refines the cavity geometry and light positions via differentiable rendering subject to geometric constraints. We calibrate the material parameters of the translucent shade material prior to optimization to improve the agreement between simulation and the fabricated result. We validate our approach by 3D-printing five lampshade designs of varying geometric complexity and demonstrate that the fabricated lampshades achieve reasonably uniform luminance.

cs.GR↗