Papers
Topics
Authors
Recent
Search
2000 character limit reached

JKO-RAG: Distributional Retrieval as Wasserstein Free-Energy Gradient Flow

Published 15 Jun 2026 in cs.IR and cs.LG | (2607.24776v1)

Abstract: RAG pipelines return a \emph{ranked list} of passages. We argue this is a mismatch: the downstream LLM conditions on a \emph{set}, and the selection problem is fundamentally geometric. We propose \jko, which frames reranking as minimising a free-energy functional F(p)=relevance+entropy+redundancyF(p)=\text{relevance}+\text{entropy}+\text{redundancy} under Wasserstein-2 gradient flow via the Jordan--Kinderlehrer--Otto proximal scheme. The ground metric Cij=(1−cos⁡⟨zi,zj⟩)2C_{ij}=(1-\cos\langle z_i,z_j\rangle)^2 encodes the semantic geometry of the embedding manifold. Our central contribution is a \emph{linear-response theory} explaining \emph{why} the Wasserstein geometry helps: the Wasserstein and KL retrieval maps differ only in their proximal Hessian -- dense and geometry-aware for W2W^2, diagonal and geometry-blind for KL -- and this difference damps the mass transport that query paraphrase induces. The theory yields a falsifiable prediction: the stability advantage is monotonically decreasing in step size hh. We verify this empirically via free-energy descent, frequency-resolved perturbation response, the predicted hh-dependence, and a certified-radius analysis. Four extensions are introduced: \textbf{\nmjko} (learned ground metric), \textbf{\bwjko} (W2W^2--KL interpolation), \textbf{\samjko} (2×2\times speedup), and \textbf{\dualrank} (OT dual potentials as confidence signals). Across five BEIR benchmarks, \jko\ outperforms the cross-encoder on all five; the decisive advantage is robustness -- 22--38\% more stable under paraphrase, 2×2\times fewer leaked distractors.

Authors (2)

Summary

No one has generated a summary of this paper yet.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Continue Learning

We haven't generated follow-up questions for this paper yet.