Generated Query Expansion Still Helps Strong Sparse Retrieval: A Controlled Study with SPLADE-v3

Ryan C. Barron; Cade W. Trotter; Maksim E. Eren; Kim Ø. Rasmussen; Liz D. Miller; Benjamin J. Migliori

cs.IRSep 29, 2026

Generated Query Expansion Still Helps Strong Sparse Retrieval: A Controlled Study with SPLADE-v3

Authors: Ryan C. Barron, Cade W. Trotter, Maksim E. Eren, Kim Ø. Rasmussen, Liz D. Miller, Benjamin J. Migliori

Organizations: Computational Intelligence & Modeling, Los Alamos National Laboratory, Los Alamos, New Mexico, USA. · Modeling and Observations of Earth Systems, Los Alamos National Laboratory, Los Alamos, New Mexico, USA. · Fluid Dynamics and Solid Mechanics, Los Alamos National Laboratory, Los Alamos, New Mexico, USA. · Intelligence & Systems Analysis, Los Alamos National Laboratory, Los Alamos, New Mexico, USA. · Advanced Research in Cyber Systems, Los Alamos National Laboratory, Los Alamos, New Mexico, USA.

Abstract

Scientific queries are often brief, while relevant papers use specialized vocabulary. Generated query expansion can bridge this mismatch, but earlier work suggests that its value shrinks as the underlying retriever becomes stronger. We test the four generated formats of term lists, a pseudo-document, multiple pseudo-references, and corpus-steered text all together with SPLADE-v3 on NFCorpus, TREC-COVID, and SciDocs. Every condition searches the same frozen document index and follows the same query-side integration rule and 256-dimension budget, isolating the effect of the added content. All twelve method-collection comparisons improve aggregate nDCG@10, with best relative gains of 4.81%, 8.92%, and 9.47%. Eleven remain significant after Holm correction. The gain persists in 103 of 114 interpolation settings, including every setting that assigns at least 30% of the mixture weight to the original query. Shuffled-text and non-contextual lexical-bag controls also remain above baseline in all 24 aggregate comparisons, showing that the added vocabulary carries most of the benefit. A corpus-induced typed concept graph, by contrast, produces no consistent gain, and its relation, depth, validation, random, and gating controls do not rescue it. Generated vocabulary can therefore complement a strong learned sparse retriever, provided that the original query remains strongly represented.

Figures & tables

	NFCorpus		TREC-COVID		SciDocs
Method	nDCG@10	R@100	nDCG@10	R@100	nDCG@10	R@100
BM25	0.3231	0.2457	0.5696	0.1091	0.1490	0.3477
RM3	0.3465	0.3229	0.5635	0.1168	0.1491	0.3620
TAS-B dense	0.2755	0.2487	0.3087	0.0335	0.1406	0.3222
BM25 + dense RRF	0.3342	0.2819	0.5511	0.0894	0.1690	0.3848
ColBERTv2	0.3392	0.2803	0.7100	0.1307	0.1482	0.3543

TABLE I: First-stage effectiveness, with nDCG@10 as the primary metric, R@100 as deeper coverage, and bold marking the best value in each dataset column.

Dataset	Method	nDCG@10	MRR@10	MAP@10	P@10	R@10	R@100	Cand. recall
NFCorpus	CSQE	+4.81	+5.10	+5.30	+4.45	+1.81	+6.20	+6.28
TREC-COVID	Query2doc	+8.92	+3.74	+6.48	+3.91	+6.99	+10.44	+5.11
SciDocs	CSQE	+9.47	+9.16	+12.58	+7.38	+7.39	+5.57	+3.14

TABLE II: Relative improvement (%) over SPLADE-v3 for the strongest generated nDCG@10 condition on each collection.

Dataset	Method	$\Delta$	95% CI	$p_{\mathrm{adj}}$	$d_{z}$	W/T/L
NFCorpus	CSQE	+0.0173	[+0.0106, +0.0241]	4e-05	+0.28	99/175/49
NFCorpus	HiQE , ungated	+0.0013	[-0.0007, +0.0034]	0.661	+0.07	40/246/37
TREC-COVID	Query2doc	+0.0649	[+0.0273, +0.1030]	0.00540	+0.47	34/5/11
TREC-COVID	HiQE , ungated	-0.0126	[-0.0259, +0.0007]	0.207	-0.26	16/12/22
SciDocs	CSQE	+0.0150	[+0.0111, +0.0189]	4e-05	+0.24	262/611/127
SciDocs	HiQE , ungated	+0.0006	[-0.0011, +0.0022]	1.000	+0.02	77/843/80

TABLE III: Compact paired per-query nDCG@10 analysis comparing, for each collection, the strongest aggregate generated condition and ungated HiQE with SPLADE-v3.

NFCorpus
Method	Positive $\alpha$	BOW $\Delta$	Shuffle $\Delta$
Flat LLM-QE	8/10	+.0094	+.0104
Query2doc	8/9	+.0107	+.0141
MuGI-style	9/9	+.0085	+.0104
CSQE	10/10	+.0180	+.0167
TREC-COVID

TABLE IV: Interpolation robustness and same-content controls. “Positive

\alpha

” counts tested weights above SPLADE-v3. BOW and Shuffle are absolute

\Delta

nDCG@10 at the configured

\alpha

Variant	NFCorpus	TREC-COVID	SciDocs	Mean $\Delta$
Random count-matched	$-.0001$	$-.0060$	$+.0003$	$-.0020$
Flat concepts, equal scores	$-.0048$	$-.0223$	$-.0002$	$-.0091$
No corpus validation	$+.0006$	$-.0256$	$+.0006$	$-.0081$
Equal relation weights	$+.0005$	$-.0192$	$+.0014$	$-.0058$
Parent only	$+.0002$	$-.0054$	$+.0005$	$-.0016$
Child only	$+.0006$	$-.0093$	$+.0004$	$-.0028$

TABLE V: Hierarchy ablations as absolute

\Delta

nDCG@10 relative to SPLADE-v3, with mean

\Delta

defined as the arithmetic mean across the three collections.

Explore similar work

ADORE: Iterative Query Expansion with Retrieval-Grounded Relevance Feedback
Jun 11, 2026Amin Bigdeli, Negar Arabzadeh, Radin Hamidi Rad +3Relevance Feedback
University of Waterloo · University of California, Berkeley · Mila – Quebec AI Institute +1
Query Expansion Should Be Coordinated: Dense Expands, Sparse Anchors
Aug 16, 2026Chunran ZhangDense
School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu, China
Latent Terms: Dense Retrievers Contain Trivially Extractable BM25-ready Zipfian Vocabularies
May 28, 2026Benjamin Clavié, Sean Lee, Aamir Shakir +1Retrieval Layer Retrievers
Mixedbread AI · National Institute of Informatics (NII) · University of Tsukuba

Jun 11, 2026, cs.IRJ/K move · Enter open · S save

ADORE: Iterative Query Expansion with Retrieval-Grounded Relevance Feedback

Amin Bigdeli, Negar Arabzadeh, Radin Hamidi Rad, Sajad Ebrahimi+2

University of Waterloo · University of California, Berkeley · Mila – Quebec AI Institute+1

Relevance Feedback

Open paper AskPDF arXiv

LLM-based query expansion improves retrieval by enriching the original query with additional context. Yet most methods remain generation-driven, producing plausible pseudo-documents or expansions without checking how the target corpus responds. This can introduce retrieval drift, amplify misleading vocabulary, or miss terms that distinguish relevant from non-relevant documents. We argue that effective expansion requires retrieval-grounded feedback, not just single-pass generation or unverified iteration. We introduce ADORE (ADapt, Observe, Relevance Evaluate), an iterative framework that turns retrieval outcomes into feedback for the next expansion. At each round, an LLM generates pseudo-passages, a retriever exposes the corpus response, and a relevance assessor evaluates retrieved documents against the original query. These judgments identify what to reinforce, what remains undercovered, and what to suppress. Across TREC Deep Learning, BEIR, and BRIGHT, ADORE consistently outperforms strong query expansion baselines with notable improvements across nearly all evaluation settings, improving average nDCG@10 by 24.5% over BM25 and 3.6% over the strongest prior query expansion method on BEIR, and by 122.9% over BM25 and 9.2% over the best query expansion baseline on BRIGHT. Our code and data are publicly available.

Abstract

Explore similar work

ADORE: Iterative Query Expansion with Retrieval-Grounded Relevance Feedback

Query Expansion Should Be Coordinated: Dense Expands, Sparse Anchors

Latent Terms: Dense Retrievers Contain Trivially Extractable BM25-ready Zipfian Vocabularies