---
title: "Bridge Evidence: Static Retrieval Utility Does Not Predict Causal Utility in Multi-Step Agentic Search"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.15253v1"
permalink: "https://feed.astralor.com/items/itm_33fea838d0d5223dae30adca"
published: "2026-07-16T17:48:29.000Z"
score: 79
channel: "ai"
tags: ["Agentic Search","HotpotQA","ReAct","Causal Inference","Bridge Evidence","Retrieval Utility"]
---

# Bridge Evidence: Static Retrieval Utility Does Not Predict Causal Utility in Multi-Step Agentic Search

## AI commentary

> This paper hits the critical blind spot in current RAG-Agent integration: we are still measuring dynamic processes with static metrics. For researchers working on search-augmented applications and evaluation, this work directly shakes your understanding of 'relevance' and is well worth a deep read.

## At a glance

This paper reveals that current retrieval systems evaluated on static utility (a document directly helps answer a question) break when an LM acts as a search agent. Using a ReAct agent on HotpotQA, experiments across 23,322 document observations show static RAG utility and counterfactual trajectory utility (CTU) are nearly independent (Spearman ρ=-0.026). About one third of documents read by the agent are causally load-bearing yet appear useless to a static reader—termed bridge documents. Mechanistic analysis shows bridge documents provide discriminative entities that redirect subsequent agent queries. Static relevance and causal usefulness are distinct quantities in agentic retrieval; optimizing one does not deliver the other.

## Key points

- Static RAG utility and counterfactual trajectory utility (CTU) are nearly independent across 23,322 observations (Spearman ρ=-0.026), showing traditional evaluation fails in agentic settings.
- About one third of 'bridge documents' are useless in static reading but causally critical for the agent's next queries, providing discriminative entities that redirect search.
- Discriminative entities from bridge documents appear in the agent's next query 4.02 times more often than entities from non-relevant documents (6.1% vs 1.5%).

## Full text

Retrieval systems are trained and evaluated on a static idea of usefulness: hand a document and a question to a reader model, see whether the answer improves, and score the document accordingly. The idea holds up when a document is read on its own. It breaks when a language model works as a search agent, issuing several queries and reasoning across turns, because a document can matter for what it lets the agent do next rather than for what it says about the current question. We measure that gap rather than argue it. Using a ReAct style agent over HotpotQA, we replay 1000 development questions and, for every document the agent read, delete it and re-run the rest of the trajectory from that point. Comparing the original run against its counterfactual gives a Counterfactual Trajectory Utility (CTU) score from three deltas: final answer quality, next query retrieval quality, and turn count. Crossing CTU against Static RAG Utility (SRU) over 23,322 document observations, the two are close to statistically independent (Spearman rho = -0.026). Roughly a third of the documents the agent reads are causally load bearing while looking useless to a static reader; we call these bridge documents. The pattern survives when the reader based axis is swapped for a BM25 and cross encoder proxy, giving a bridge cell of 27.2% on an evenly spread axis. A second experiment pins down the mechanism. Using the Observable Entity Relevance (OER) measure from prior work, entities that discriminate relevant from non-relevant candidates appear in the agent's next query 4.02 times more often than entities found only in non-relevant documents (6.1% vs 1.5%, n = 227,139). A bridge document earns its keep by handing the agent a discriminative entity that redirects the search. Static relevance and causal usefulness are different quantities in agentic retrieval, and optimizing the first does not deliver the second.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_33fea838d0d5223dae30adca