AB-RAG: Adaptive Budgeted Retrieval-Augmented Generation for Reliable Question Answering

Retrieval Granularity as Evidence Design in Small-Model RAG Question Answering: A Diagnostic HotpotQA Study

Results align with a diagnostic perspective on chunking: using evidence at a task-appropriate level of granularity can improve grounding, auditability, and answer quality, but the observed patterns should be interpreted within the HotpotQA distractor setting, fixed generator, and tested context budgets.

Evaluating RAG approaches for Question Answering

Know Before You Fetch: Calibrated Retrieval-Budget Allocation for Retrieval-Augmented Generation

Retrieval-augmented generation (RAG) typically retrieves a fixed number of passages for every query. This is wasteful when the reader already knows the answer, and it can be harmful when irrelevant or partially relevant passages distract the reader. We formulate adaptive RAG as calibrated retrieval-budget allocation: given a query, decide whether to answer closed-book, retrieve a compact context (k=1), retrieve a full context (k=5), or abstain. The contribution is a probability interface rather than a new raw uncertainty signal. We calibrate sequence log-probability and prefix-logit uncertainty signals into probabilities of correctness, then use these probabilities for graded context selection, selective abstention, and explicit latency/token trade-offs. Across core QA experiments on TriviaQA, Natural Questions, and MS MARCO, with auxiliary PopQA motivation and Qwen/Llama family checks, diagnostic out-of-fold calibration improves probability quality dramatically: for sequence log-probability, ECE drops from 0.275 to 0.062 on TriviaQA, 0.643 to 0.009 on NQ, and 0.711 to 0.031 on MS MARCO. Graded retrieval improves full-context and passage-budget frontiers for both our signal and TARG-style prefix entropy/margin, while retrieval-call AUC remains essentially tied with binary gating because k=1 is still a retrieval call. Held-out train/validation/test threshold experiments report deployable operating points. At matched-accuracy frontier operating points, a measured cost model reveals that gating is not universally faster: it increases latency by about 27% on Qwen3-8B but saves about 8% on Qwen3-32B. These results support a nuanced view of adaptive RAG: calibrated confidence is best understood as a reusable interface for allocating retrieval budget under task and system constraints.

Testing Retrieval-Augmented Generation Systems with Chunk Coverage

Retrieval-Augmented Generation (RAG)-based systems\footnote{For brevity, RAG-based systems are referred to as RAG systems throughout this paper.} are increasingly deployed in high-stakes settings where correct behaviour depends not only on the language model but also on the retrieval component that selects external documents at inference time. While existing RAG evaluation metrics assess retrieval and generation quality on a per-query basis, typically relying on query-level test oracles such as reference answers or relevance annotations, they provide limited insight into whether a test suite adequately exercises the retrieval behaviour of the system as a whole. In this paper, we introduce Chunk Coverage (CC), an oracle-independent test adequacy criterion for testing the retrieval component of RAG systems. CC measures the fraction of corpus chunks that are retrieved at least once across a test suite, providing a structural view of which parts of the retrieval space have been exercised. We further show how CC can be used to guide test selection and generation by prioritising queries that expand coverage of previously unexercised retrieval regions. We evaluate CC on clinical and financial RAG system scenarios. CC-guided testing reaches 50% of attainable coverage 1.7x faster than random selection and 4.2x faster than redundancy-biased strategies. Moreover, CC improves fault detection effectiveness (APFD) by 10% to 25% over random, indicating earlier discovery of distinct retrieval faults. These results show that CC captures retrieval diversity relevant to effective testing without requiring test oracles.

QUBO-Optimized Evidence Selection for Retrieval-Augmented Question Answering with Unconventional Solvers

Retrieval-augmented question answering depends on selecting evidence passages that jointly support answer generation. However, many RAG pipelines rely on top-\(k\) ranking, where passages are selected mainly by individual relevance scores, even though multi-hop questions often require complementary evidence satisfying multiple information requirements. Recent LLM-based selectors address this by treating retrieval as set selection, but using an LLM for this intermediate stage can be costly and difficult to scale. In this work, we formulate evidence selection as a Quadratic Unconstrained Binary Optimization (QUBO) problem. Given a question, candidate passages, and decomposed information requirements, our method constructs an energy function that balances relevance, requirement coverage, support strength, redundancy, complementarity, and compactness. Low-energy solutions correspond to compact evidence subsets that cover the needed requirements while avoiding unnecessary or repetitive context. The selected passages are then passed to a downstream language model for answer generation, separating combinatorial evidence selection from semantic answer generation. We evaluate the proposed QUBO selector on HotpotQA and compare it with LLM-based set selectors and non-LLM baselines including BM25, relevance top-\(k\), maximal marginal relevance, hybrid lexical--semantic ranking, greedy coverage, and random selection. The QUBO selector achieves competitive exact-match and token-F1 performance relative to LLM-based selectors while providing a solver-compatible formulation for structured evidence selection. These results suggest that multi-hop evidence selection can be cast as discrete optimization, opening a path toward RAG pipelines where LLMs are reserved for semantic processing and answer generation, while context selection is handled by Ising/QUBO-compatible solvers.

Effectiveness of Sparse Retrieval for Automated Fact-Checking

As fact-checking pipelines grow increasingly complex, a critical question remains unanswered: does retrieval matter more than reasoning? We investigate this question on the AVeriTeC 2.0 Shared Task, a leading and highly competitive benchmark for automated fact-checking that requires systems to verify real-world claims against open-domain web evidence. On this benchmark, we compare dense retrieval (FAISS), learned sparse retrieval (SPLADE), late interaction (ColBERT), and contextual embeddings across two architecturally distinct systems. SPLADE consistently outperforms all alternatives, improving new AVeriTeC scores by up to 86% on the development set and up to 27% on the hidden test set, where it enables a lower-ranked system to claim the top leaderboard position. Its advantage stems from vocabulary-space representations that bridge lexical gaps, preserve fine-grained numerical and entity matches, and remain robust to noisy web evidence, capabilities that are difficult to achieve with dense retrieval without additional techniques such as query expansion, reranking, or entity-aware training. Strikingly, two systems of vastly different complexity, one consolidating all reasoning into a single LLM call and the other orchestrating five models across seven tasks, converge to virtually identical performance once equipped with SPLADE. This demonstrates that retrieval quality is a major bottleneck in these systems. Beyond accuracy, SPLADE reduces preprocessing time by an order of magnitude while allowing both pipelines to share a single index. Our findings establish learned sparse retrieval as a highly impactful design choice in automated fact-checking, with retrieval quality serving as a critical determinant of end-to-end performance in the studied systems.