STEC: Evidence Compression for Deep Search in Open-domain Multi-Hop QA
2026-07-12 • Artificial Intelligence
Artificial IntelligenceComputation and Language
AI summaryⓘ
The authors study how to improve answering complex questions that need multiple steps and lots of information gathering. They point out that picking the best final answer is hard because different search paths can give conflicting or incomplete information. To fix this, they introduce a method called STEC that groups evidence by each answer to create a clearer comparison before choosing the final answer. Their tests show STEC works better than other methods, and breaking down evidence by answer is helpful for picking the right one.
open-domain question answeringmulti-hop reasoninglarge language modelssearch trajectoriesanswer verificationevidence compressionretrievalfinal answer selectionbenchmark evaluation
Authors
Xinkang Li, Rong Jiang, Xin Song, Ye Wang, Yue Han, Changjian Li
Abstract
In open-domain multi-hop question answering (QA), LLM-based search agents offer a promising approach to knowledge-intensive QA by combining retrieval with reasoning. Existing methods mainly improve open-domain multi-hop QA through reasoning paradigms, retrieval interaction, and search strategy optimization. However, using multiple search trajectories introduces a challenging final answer selection problem. Different trajectories may support different candidates, and the retrieved information can be heterogeneous, redundant, incomplete, or conflicting. Directly comparing raw trajectories exposes the verifier to noisy and unaligned content, while comparing answer strings ignores the evidence supporting each candidate, making reliable final selection difficult. To address this challenge, we propose STEC, an evidence compression framework for final answer selection in multi-hop QA. STEC selects the final answer from the existing candidate set through two mechanisms: (1) Answer-Level Evidence Compression, which groups trajectories by normalized answer identity and converts each answer group into a candidate-specific evidence representation; and (2) Evidence-Guided Answer Verification, which compares these representations and selects the final answer from the candidate set. The design shifts final selection from raw trajectory comparison to candidate-level evidence comparison. We evaluate STEC on four open-domain multi-hop QA benchmarks against representative baselines. Experimental results show that STEC performs best overall among the compared methods, and ablation results provide evidence that answer-level evidence compression contributes to final answer selection.