Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Entropic Claim Resolution: Uncertainty-Driven Evidence Selection for RAG

Davide Di Gioia

Abstract

Current Retrieval-Augmented Generation (RAG) systems predominantly rely on relevance-based dense retrieval, sequentially fetching documents to maximize semantic similarity with the query. However, in knowledge-intensive and real-world scenarios characterized by conflicting evidence or fundamental query ambiguity, relevance alone is insufficient for resolving epistemic uncertainty. We introduce Entropic Claim Resolution (ECR), a novel inference-time algorithm that reframes RAG reasoning as entropy minimization over competing semantic answer hypotheses. Unlike action-driven agentic frameworks (e.g., ReAct) or fixed-pipeline RAG architectures, ECR sequentially selects atomic evidence claims by maximizing Expected Entropy Reduction (EER), a decision-theoretic criterion for the value of information. The process dynamically terminates when the system reaches a mathematically defined state of epistemic sufficiency (H <= epsilon, subject to epistemic coherence). We integrate ECR into a production-grade multi-strategy retrieval pipeline (CSGR++) and analyze its theoretical properties. Our framework provides a rigorous foundation for uncertainty-aware evidence selection, shifting the paradigm from retrieving what is most relevant to retrieving what is most discriminative.

Entropic Claim Resolution: Uncertainty-Driven Evidence Selection for RAG

Abstract

Current Retrieval-Augmented Generation (RAG) systems predominantly rely on relevance-based dense retrieval, sequentially fetching documents to maximize semantic similarity with the query. However, in knowledge-intensive and real-world scenarios characterized by conflicting evidence or fundamental query ambiguity, relevance alone is insufficient for resolving epistemic uncertainty. We introduce Entropic Claim Resolution (ECR), a novel inference-time algorithm that reframes RAG reasoning as entropy minimization over competing semantic answer hypotheses. Unlike action-driven agentic frameworks (e.g., ReAct) or fixed-pipeline RAG architectures, ECR sequentially selects atomic evidence claims by maximizing Expected Entropy Reduction (EER), a decision-theoretic criterion for the value of information. The process dynamically terminates when the system reaches a mathematically defined state of epistemic sufficiency (H <= epsilon, subject to epistemic coherence). We integrate ECR into a production-grade multi-strategy retrieval pipeline (CSGR++) and analyze its theoretical properties. Our framework provides a rigorous foundation for uncertainty-aware evidence selection, shifting the paradigm from retrieving what is most relevant to retrieving what is most discriminative.

Paper Structure

This paper contains 57 sections, 1 theorem, 11 equations, 4 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Dense, Graph-Augmented, and Agentic Retrieval
  4. Uncertainty Quantification (UQ) in RAG
  5. Entropy-Aware Context Management
  6. Claim-Level Verification and Value of Information
  7. Methodology: Entropic Claim Resolution (ECR)
  8. Problem Formulation
  9. Objective Function: Answer Entropy
  10. Expected Entropy Reduction (EER) and Selection Policy
  11. Implementation-Level EER Proxy.
  12. Design choice of the EER proxy.
  13. Coherence-aware selection.
  14. Contradiction-aware coherence term.
  15. Bayesian Posteriors and Epistemic Sufficiency
  16. ...and 42 more sections

Key Result

Theorem 1

For any finite candidate set $\mathcal{C}_{cand}$, ECR terminates after at most $\min(T, |\mathcal{C}_{cand}|)$ claim evaluations. Moreover, if there exists a constant $\delta>0$ such that at each iteration the selected claim satisfies $\mathbb{E}[H_{t-1}-H_t] \ge \delta$ whenever $H_{t-1}>\epsilon$

Figures (4)

  • Figure 1: Pseudo-code for ECR without external algorithm packages.
  • Figure 2: System overview: Entropic Claim Resolution (ECR) operates as an inference-time controller between competitive retrieval and answer synthesis. Given a retrieved claim set, ECR sequentially selects evidence to minimize hypothesis entropy and terminates when epistemic sufficiency is reached.
  • Figure 3: Schematic entropy trajectories consistent with the measured endpoints: ECR reaches $H\le\epsilon$ quickly, whereas relevance-only and random baselines typically remain above $\epsilon$ at matched claim budgets.
  • Figure A.1: Ambiguity exposure as a function of the coherence weight $\lambda_{\mathrm{conflict}}$ under structured contradiction injection. Empirically, ambiguity exposure exhibits a sharp phase transition: for $\alpha=0.5$, exposure jumps from 0 to 1 for any tested $\lambda>0$, while remaining 0 for $\alpha \le 0.3$ across all tested settings.

Theorems & Definitions (2)

  • Theorem 1: Termination and Budget Bound
  • proof