Train, Retrieve, or Both? A Four-Arm Head-to-Head for Correct Statutory Citation on the Ontario Residential Tenancies Act
arXiv:2606.20359v1 [cs.LG] 18 Jun 2026
Ali Asaria Transformer Lab
Tony Salomone Transformer Lab
Deep Gandhi∗ Transformer Lab
Abstract Self-represented tenants, landlords, and help-desk staff need to be pointed at the provision of law that actually governs a question, with a correct statutory citation. We study this task on the Ontario Residential Tenancies Act, 2006 (RTA) and its core regulation, asking the operator’s question empirically: is fine-tuning enough, or is hybrid retrieval needed? We run a four-arm head-to-head on Qwen2.5-7B-Instruct (base zero-shot, LoRA SFT-only, RAG-only, and an SFT+RAG hybrid), scored on citation exact-match (section+subsection) over a small, human-verification-pending real eval set. The base model cannot cite the RTA and SFT-only mis-recalls sections; retrieval is essential and drives hallucination to zero by construction; and the SFT+RAG hybrid scores highest at 0.481 exact-match with zero hallucinated citations. Its edge comes from SFT making provision selection more robust to the higher-recall candidate sets that hurt zero-shot RAG. Notably, this cheap bge-small hybrid matches or beats a pipeline built on bigger, specialized retrieval models (a larger embedder and a cross-encoder reranker), and a larger/improved training set does not help either: strong statutory-citation performance here does not require specialized retrieval models or more data. The artifact zeroes hallucination and clears the lift-over-base bar but does not reach the aspirational 0.70 exact-match target. All results are on a small, human-verification-pending real eval set and are reported as preliminary.
1
Introduction
Ontario’s residential-tenancy law is governed by the Residential Tenancies Act, 2006 (RTA, S.O. 2006, c. 17) and O. Reg. 516/06 (General). The people who most need it (self-represented tenants and landlords, paralegals, and tenant-help-desk staff) rarely need fluent prose; they need to be pointed at the provision that governs, with a correct citation (the right Act/regulation, section, and subsection). We therefore treat the headline quality target as the citation score rather than answer fluency: the unit of prediction is a single natural-language tenancy question (e.g. “How much notice must a landlord give to enter my unit?”) mapped to a short answer plus one or more statutory citations (e.g. RTA s. 26(1)). This is decision-support, not legal advice. The operator posed a concrete, falsifiable question: is fine-tuning enough, or do we need hybrid RAG? Rather than assume an answer, we measure it. We build a citation index and a question→citation dataset (no public Ontario-RTA set exists) and run a four-arm head-to-head on ∗
Corresponding author: [email protected]
1
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
the same held-out citation metric: the base instruct model zero-shot, LoRA supervised fine-tuning on synthetic question→citation pairs, RAG over a chunked statute index, and the SFT+RAG hybrid. Beyond a single-GPU LoRA budget, the stated design constraints include interactive latency, a pinned consolidation date, and using no personal data as model input. Contributions. 1. A measured four-arm answer to “train vs. retrieve”: base cannot cite (0.00, 81% hallucinated), SFT-only mis-recalls sections (0.148), RAG is essential (0.44, zero hallucination), and the hybrid scores highest (0.481), giving the ordering base ≪ SFT-only ≪ retrieval-bearing arms; see §5. 2. An SFT×k effect: more retrieved candidates hurt zero-shot RAG (0.44→0.37 at k=10) but help the SFT model (0.333→0.481), so the hybrid gains selection robustness to higher-recall candidate sets; see §5. 3. An efficiency result: the cheap bge-small hybrid matches or beats a pipeline built on bigger, specialized retrieval models (a larger embedder + a cross-encoder reranker, 0.37), and a larger/improved training set does not improve over the original (0.444); see §5. 4. An honest accounting: zero hallucinated citations by construction and a +0.48 exact-match lift, but the 0.70 target is not met on a 27-item, human-verification-pending eval set; see §6.
2
Related Work
Generating verifiable citations. Teaching attribution directly is more effective than passively tagging text. Huang et al. [1] show that bidirectional source↔fact training with fact-variation augmentation lifts citation precision substantially, while verbatim token replay gives no benefit and more replay epochs can hurt generalization, which directly motivates our paraphrase-based synthetic generation and group-aware split. Pezeshkpour and Hruschka [2] find continual pre-training on raw documents has only a marginal effect, whereas reformatting to atomic facts yields large exact-match gains, supporting SFT on (question→provision-id) pairs over dumping statute prose. Xu et al. [3] self-generate QA from an unlabeled corpus with a retrieval round-trip filter, and Yasuno [4] scale graph-derived synthetic data; both inform our synthetic-from-statute recipe. Attribution methods such as TRACE [5] and judgement-citation retrieval by contextual similarity [6] frame citation as a retrieval/grounding problem rather than free generation. Retrieval for statute. Statute retrieval is semantics-dominant yet benefits from a lexical anchor: Paul et al. [7] report optimal fusion heavily weighted toward the dense retriever, but Arafat [8] and Nigam et al. [9] show BM25 is essential to anchor exact identifiers and defined terms, with a robust fusion weight that transfers without per-corpus tuning, so we adopt a hybrid BM25+dense default. Second-stage re-ranking [7] or cross-encoder rerankers [9] can lift retrieval F1; we test the reranker as an ablation. Kabir et al. [10] use relevance-check/query-refinement loops and low decode temperature, and bounded-size chunking with overlap is standard [11, 12]; we instead chunk on statutory section boundaries as a design choice, one provision per chunk. Crucially, Yasuno [4] report that bad retrieval actively hurts (over-retrieval misleads generation), and Hsu et al. [13] note multi-stage pipelines compound failures; both foreshadow our finding that heavier retrieval helps neither arm. 2
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
Anti-hallucination guardrails. Arafat [8] use mechanical citation verification (accept a citation only if it exists in the pinned corpus) to reach high accuracy with zero hallucinations; Noël et al. [14] reduce structural hallucination detection to a cheap entity-grounding existence check but warn it collapses on short, sparse inputs, favoring a deterministic existence check for single subsections. Hsu et al. [13] contrast large/small models to verify grounding, and Komorowski et al. [15] explore decode-time attribution; we adopt the cheap existence-check guardrail as a hard gate. Reality checks abound: RAG does not eliminate hallucination [16], commercial legal RAG tools still hallucinate 17–33% [14], and retrieval can game cheap metrics like ROUGE without improving judged quality [16]. LoRA and positioning. Our central positioning is simple: Yasuno [4] ran three of our four arms and explicitly left SFT+RAG as future “Case D”, and this work fills that gap on a statutory-citation exact-match metric, with a deterministic hallucination check that the surveyed legal papers measured only qualitatively. On the training side, Rathore et al. [17] show LoRA usually matches or beats full SFT while forgetting less, and that fine-tuning can erode a strong base, arguing to keep the zero-shot base as a real baseline (which we do) and to prefer LoRA.
3
Method
Task and label. For a question q we predict a short answer and a set of citations Ĉ = {(instrument, section, subsection)} against a gold set C. Correctness is section+subsection match; section-only matches earn partial credit (0.5); the instrument must match. Citations to provisions absent from the pinned consolidation count as hallucinations. Corpus and index. We parse the e-Laws consolidation of the RTA (297 sections, 895 subsections, 20 Parts) and O. Reg. 516/06 (61 sections, 166 subsections) into a citation inventory of 358 provisions, each a single structure-aware RAG chunk. The inventory doubles as the existence-check whitelist for the hallucination metric. Arms. The base arm is the instruct model zero-shot. The SFT-only arm is a LoRA SFT on synthetic question→citation pairs (rank r=32, α=2r, lr 2e−4, 400 steps; q/k/v/o projections), following the LoRA guidance of Rathore et al. [17]. The RAG-only arm is the base model with hybrid BM25+bge-small retrieval, prompted to cite only from the retrieved context (a cite-fromcontext guardrail that is effectively an existence check). We note up front that this guardrail and the hallucination metric share the same pinned inventory, so a 0.00 hallucination rate for the retrieval-bearing arms is largely guaranteed by construction rather than learned; we treat it as a design property, not an emergent result. The SFT+RAG hybrid is the SFT model trained to select the correct citation from retrieved candidates, evaluated with the same retrieval. We sweep retrieval depth k and retrieval strength (cheap bge-small vs. bge-large + cross-encoder reranker). Objective. The optimization target is the citation set, scored as exact-match with section-only partial credit and multi-citation F1. Two reporting conventions matter. First, because section-only matches earn 0.5, the “exact-match” figures we report are a partial-credit citation score in [0, 1], not a strict 0/1 rate, so we keep the name “exact-match” for brevity but it is partial-credit-weighted. Second, F1 is computed per item over the predicted vs. gold citation sets (multi-citation precision and 3
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
recall, then their harmonic mean) and averaged across items; it can exceed exact-match because it gives proportional credit for partially-correct citation sets. Conceptually, an arm’s quality combines retrieval supply and selection: ExactMatch ≤
× P (select gold | retrieved),
recall@k | {z }
retrieval supplies gold
|
{z
SFT improves this
(1)
}
which frames the central finding: SFT raises the selection term while RAG raises the supply term, and the hybrid benefits from both.
4
Experimental Setup
Data. No public Ontario-RTA question→citation dataset exists, so s4 builds one. Training (synthetic). An LLM paraphrases each section/subsection into natural questions with the provision as gold: 2,148 pairs, split leakage-safely (question-level bigram-Jaccard dedup + a 15% unseensection slice) into train 1,473 / synth_test 345 / unseen 330. Held-out (real, headline). A small set mined from tenant-law sites with explicit RTA section references: 27 source-cited, in-inventory items, each flagged human-verification-pending. Splits are group-aware by section/Part so the model cannot win by memorizing question→citation maps; the real set is fully source-disjoint. Statute text is the e-Laws consolidation (Government of Ontario, Crown copyright). Protocol. The base model is Qwen2.5-7B-Instruct (operator: Qwen-only; Llama-3.1-8B was not evaluated, HF-gated). Metrics: citation exact-match (instrument+section+subsection) with section-only partial credit, multi-citation precision/recall/F1, a deterministic hallucinated-citation rate (existence check against the pinned inventory), and recall@k for the RAG arms to separate retrieval from generation failure. The scorer is deterministic and reused across arms. A harness control (perfect predictions on the 27 items → exact-match 1.0, hallucination 0.0) confirms the scorer and gold are mechanically correct: it validates the measurement instrument, not the statistical reliability of the model numbers, which is bounded by n=27. All arms are single-run point estimates (no seed replication). Total compute was 4.43 GPU-hours against a planned 44 (Lambda H100 for training, SkyPilot RTX 3090 for eval).
5
Results
Table 1 reports the final model against the overview’s success-criteria gates, and Table 2 the full arm/ablation grid. All numbers are single-run point estimates on the 27-item real eval (humanverification-pending) unless noted, so they should be read as preliminary. On 27 items a single correct/incorrect prediction moves a rate by ∼0.037, so we draw inferences only at the level of large, qualitatively-robust gaps (base vs. SFT-only vs. the retrieval-bearing arms); differences of one to three items among the retrieval-bearing cells (e.g. 0.481 vs. 0.44, or the k and retrieval-strength sweeps) are within noise and are reported as suggestive, not significant. We also flag a selection caveat: the final configuration (k=10) was chosen as the best-scoring cell on this same 27-item set, so the headline 0.481 is a max-over-configurations and is optimistically biased (see §6). Train vs. retrieve (Fig. 1). The base model cannot cite the RTA: 0.00 exact-match with 81% of citations hallucinated. SFT-only teaches the citation format and cuts the hallucination rate from 4
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
Table 1: Final model (SFT+RAG hybrid, k=10) vs. success-criteria gates. Three of four gates are met; the absolute exact-match target is not. Metric
Result
Target
Verdict
Citation exact-match (sec+subsec) Hallucinated-citation rate Lift over zero-shot base Evidenced train-vs-retrieve answer
0.481 0.00 +0.48 yes
≥ 0.70 ≤ 0.05 ≥ 0.15 —
not met met met met
Table 2: Four-arm head-to-head and ablations on the 27-item real eval (single-run point estimates). SFT-only learns format but mis-recalls sections; RAG is essential and zeroes hallucination by construction; the hybrid at k=10 scores highest, though gaps among the retrieval-bearing rows are within noise at n=27. The lower block shows that swapping the cheap bge-small retriever for a bigger, specialized pipeline (a larger embedder and a cross-encoder reranker) does not help either arm (0.37 vs. the cheap hybrid’s 0.481), and does not raise recall: the efficiency result. F1 was not computed for the light-SFT and specialized-retrieval ablations (shown as “—”). Arm
Config
k
recall@k
Exact
F1
base zero-shot SFT-only
— LoRA r32/400
— —
— —
0.00 0.148
0.00 0.167
Cheap retrieval (bge-small hybrid BM25+dense) RAG-only k=5 5 0.815 RAG-only k=10 10 0.889 hybrid r32/400, k=5 5 0.815 hybrid (light) r16/150, k=5 5 0.815 hybrid (final) r32/400, k=10 10 0.889
0.44 0.37 0.333 0.333 0.481
0.52 0.46 0.43 — 0.574
Bigger, specialized retrieval (bge-large + cross-encoder reranker) RAG-only k=10 10 0.852 0.37 — hybrid r32/400, k=10 10 0.852 0.37 —
0.81 to 0.148, but mis-recalls exact sections, reaching only 0.148 exact-match (the two 0.148 figures are distinct quantities that coincide), consistent with exact section numbers being a memorization weak spot for ∼8B models [4]. RAG is the decisive lever: RAG-only reaches 0.44 and drives hallucination to 0.00, which (as noted in §5) follows by construction from the cite-from-context existence check rather than being learned. The hybrid scores highest at 0.481 exact-match (0.574 F1, 0.00 hallucination), though its ∼0.04 (≈1-item) margin over RAG-only is within noise at n=27. The qualitatively robust finding is the large step from the non-retrieval arms (≤0.148) to the retrieval-bearing arms (≥0.37), not the ordering within the latter. A candidate-depth (k) pattern (hypothesis). The hybrid’s apparent advantage is conditional on candidate depth. More candidates appear to hurt zero-shot RAG (0.44 at k=5 →0.37 at k=10, even as recall rises 0.815→0.889), as if distractors mislead selection; but the same depth helps the SFT model (0.333 at k=5 →0.481 at k=10). Our reading is that SFT teaches more robust selection from a noisier, higher-recall candidate set. We stress two limits on this claim. First, the 5
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
Figure 1: Citation exact-match rises monotonically across arms (base 0.00, SFT-only 0.148, RAGonly 0.44, SFT+RAG hybrid 0.481), but the best arm still falls short of the 0.70 target (dashed). Real eval, n=27, human-verification-pending. deltas are 1–4 items on n=27, so the difference-in-differences “interaction” is under-powered and we offer it as a hypothesis, not an established effect. Second, it is confounded: recall also rises with k (0.815→0.889), so the hybrid’s k=10 gain could be additional retrieval supply rather than better selection; the recall@k × P (select) decomposition in Eq. (1) is motivating, but we did not measure the conditional selection term P (select | retrieved) directly, so supply and selection cannot be cleanly separated here. A lighter SFT (r16/150) also scores 0.333 at k=5, which is consistent with (but, being two single-run points, does not establish) insensitivity to LoRA strength. Efficiency: no measurable benefit from heavier retrieval (Fig. 2). The cheap bge-small hybrid (0.481) is not beaten by a heavier retriever (bge-large + cross-encoder reranker), which scores 0.37 for RAG-only and 0.37 for the hybrid; recall is also no higher (0.889→0.852). The observed ordering is hybrid k=10 (0.481) ≥ RAG-only best (0.44) ≥ all big-retrieval cells (0.37). We emphasize that these cells span ∼3 items on n=27 with overlapping uncertainty: the safe reading is that heavier, specialized retrieval buys no measurable improvement on this short-statute corpus (a useful negative result for a latency- and cost-constrained deployment), not that the small retriever is provably superior. Separately, an improved/larger training set (2,645 pairs, 63% subsection-targeted) did not improve over the original, reaching 0.444 with 0.037 hallucination (again a ∼1-item difference); we read the data lever as offering no easy further gain here, not as exhausted. Subgroups and errors. The real slice (0.481) is, surprisingly, higher than the synthetic-test slice (0.327), the opposite of the usual synthetic-overfit worry, because real questions cluster on a few high-frequency provisions the retriever handles well, whereas synthetic questions span the long tail. Per-RTA-Part subgroups are not statistically powered at n=27 and are deferred. The dominant error mode is retrieval misses (~11%; recall@10 = 0.889): when the governing section is not in the
6
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
Figure 2: Efficiency 2×2: a cheap bge-small hybrid (0.481) is the highest-scoring cell; a heavier bge-large+reranker retriever does not improve either RAG-only (0.37) or the hybrid (0.37). On a 27-item set these cells carry overlapping uncertainty, so the reading is that heavier retrieval buys no measurable benefit on this corpus, not that the small retriever is provably better. top-10 the model guesses a topically adjacent provision (hallucination stays 0 because the guess is a real section). Secondary modes are mis-selection among retrieved candidates and subsection drop (section right, subsection missing → 0.5 partial credit), which accounts for much of the gap between F1 (0.574) and exact-match (0.481).
6
Discussion and Limitations
The results give the operator a clear answer at the level the data can support: training alone is insufficient and retrieval is essential. The hybrid is the highest-scoring arm and we read its edge as more robust selection from higher-recall candidate sets, but at n=27 its margin over RAG-only is within noise, so we do not claim the hybrid is provably best. The system passes three of four gates: it zeroes hallucinated citations (by construction) and clears the lift bar by a wide margin (0.481 vs. 0.00, target ≥ 0.15). A useful efficiency finding is that heavier retrieval and more training data both buy no measurable improvement on this short-statute corpus, which matters for a latency- and cost-constrained deployment even though our sample cannot prove the cheap retriever superior. Honestly, the headline gate is not met: at 0.481 exact-match the system is well below the aspirational 0.70 target. Three factors bound the gap: a retrieval-recall ceiling around 0.89 (the dominant error mode); subsection-level difficulty (the gap between F1 and exact-match); and a tiny, human-verification-pending eval set of 27 items. On generalization to untrained provisions, the only configuration we ran on the unseen-section slice was the improved-data retrain variant (not the headline model), which scored 0.31 exact-match on a held-out slice of 330 items whose sections were never seen in training; it degrades but does not collapse, which suggests retrieval can supply untrained provisions, though this evidence is from a different model than the headline and should
7
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
be read with that caveat. Calibration is not applicable to v1: the model emits a citation string with no confidence, so Brier and ECE are undefined, and it remains future work. 6.1
Threats to Validity
Construct and statistical power. The headline metric rides on a 27-item, human-verificationpending real eval set. A harness control confirms the scorer and gold are mechanically correct, but it does not address statistical reliability: with n=27 and single-run, no-confidence-interval estimates, one item shifts a rate by ∼0.037, so the fine-grained ordering among the retrieval-bearing arms (the 0.481 vs. 0.44 win, the candidate depth pattern, and the efficiency 2×2) is within noise and is reported as suggestive only. Two further cautions apply. First, the final configuration was selected on this same eval set, so the headline 0.481 is a max-over-configurations and is optimistically biased. Second, the favorable real-vs-synthetic gap (0.481 vs. 0.327 on the larger n=150 synthetic slice) suggests the real questions cluster on common, high-frequency provisions, so the headline likely overstates performance on the long tail; the better-powered synthetic number is arguably the more representative estimate for arbitrary questions. The remedy for all of these, and the precondition for any claim of statistical certainty, is a larger, topic-balanced, human-verified eval set, which is our top future-work item. Until then all real-set numbers should be read as preliminary. External. Training is synthetic-heavy (statute-derived paraphrases) and the study is singlejurisdiction, single-statute (Ontario RTA + one regulation) on a single base model (Qwen2.5-7B; Llama-3.1-8B not evaluated); legal-NLP components are known not to port across jurisdictions [9], so generalization beyond this setting is unproven. Internal. A retrieval-recall ceiling (~0.89) caps achievable exact-match independent of the generator; leakage is mitigated by group-aware splits and an unseen-section slice, but the eval set’s clustering on high-frequency provisions is a confound for the headline number.
7
Conclusion and Future Work
On the task of citing the correct Ontario RTA provision, a four-arm head-to-head on Qwen2.5-7B shows that fine-tuning alone is not enough and retrieval is essential. The SFT+RAG hybrid is the highest-scoring arm (0.481 exact-match, 0.574 F1, zero hallucination by construction), which we read as SFT making provision selection more robust to the higher-recall candidate sets that hurt zero-shot RAG, though at n=27 this margin is within noise rather than a proven win; on the same data, a heavier retriever and more training data both buy no measurable improvement. The artifact does not reach the 0.70 exact-match target on a 27-item, human-verification-pending eval set. Ranked next steps: (1) a larger, topic-balanced, human-verified real eval set, which is the precondition for treating any fine-grained ranking as statistically reliable; (2) a stronger retriever or reranker tuned for short statute text, attacking the dominant retrieval-miss error mode; and (3) subsection-level modeling and constrained decoding to close the gap between F1 and exact-match, plus citation-confidence calibration as a stretch goal.1 1
All artifacts (code, data, the synthetic question→citation dataset, and the trained LoRA adapter) are available from the authors on request. Statute text is from Ontario e-Laws (Government of Ontario, Crown copyright; reproducible under the King’s Printer terms, confirm attribution before redistribution). Outputs are decision-support, not legal advice.
8
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
References [1] Yukun Huang, Sanxing Chen, Jian Pei, Manzil Zaheer, and Bhuwan Dhingra. Cite Pretrain: Retrieval-free knowledge attribution for large language models. 2025. URL https://arxiv. org/abs/2506.17585. [2] Pouya Pezeshkpour and Estevam Hruschka. Learning Beyond the Surface: How far can continual pre-training with LoRA enhance LLMs’ domain-specific insight learning? 2025. URL https://arxiv.org/abs/2501.17840. [3] Ran Xu, Hui Liu, Sreyashi Nag, Zhenwei Dai, Yaochen Xie, Xianfeng Tang, Chen Luo, Yang Li, Joyce C. Ho, Carl Yang, and Qi He. SimRAG: Self-improving retrieval-augmented generation for adapting large language models to specialized domains. 2024. URL https: //arxiv.org/abs/2410.17952. [4] Takato Yasuno. Suppressing Domain-Specific Hallucination in Construction LLMs: A knowledge graph foundation for GraphRAG and QLoRA on river and sediment control technical standards. 2026. URL https://arxiv.org/abs/2603.13307. [5] Cheng Wang, Xinyang Lu, See-Kiong Ng, and Bryan Kian Hsiang Low. TRACE: TRansformerbased attribution using contrastive embeddings in LLMs. 2024. URL https://arxiv.org/ abs/2407.04981. [6] Akshat Mohan Dasula, Hrushitha Tigulla, and Preethika Bhukya. Judgement Citation Retrieval using Contextual Similarity. 2024. URL https://arxiv.org/abs/2406.01609. [7] Shounak Paul, Dhananjay Ghumare, Pawan Goyal, Saptarshi Ghosh, and Ashutosh Modi. IL-PCSR: Legal corpus for prior case and statute retrieval. 2025. URL https://arxiv.org/ abs/2511.00268. [8] Jahidul Arafat. Citation-Grounded Code Comprehension: Preventing LLM hallucination through hybrid retrieval and graph-augmented context. 2025. URL https://arxiv.org/abs/ 2512.12117. [9] Shubham Kumar Nigam, Tanmay Dubey, Noel Shallum, and Arnab Bhattacharya. Segment First, Retrieve Better: Realistic legal search via rhetorical role-based queries. 2025. URL https://arxiv.org/abs/2508.00679. [10] Muhammad Rafsan Kabir, Rafeed Mohammad Sultan, Fuad Rahman, Mohammad Ruhul Amin, Sifat Momen, Nabeel Mohammed, and Shafin Rahman. LegalRAG: A hybrid RAG system for multilingual legal information retrieval. 2025. URL https://arxiv.org/abs/2504.16121. [11] Dnyanesh Panchal, Aaryan Gole, Vaibhav Narute, and Raunak Joshi. LawPal: A retrieval augmented generation based system for enhanced legal accessibility in india. 2025. URL https://arxiv.org/abs/2502.16573. [12] Anuraj Maurya. Scaling Legal AI: Benchmarking mamba and transformers for statutory classification and case law retrieval. 2025. URL https://arxiv.org/abs/2509.00141.
9
Train, Retrieve, or Both? Statutory Citation for the Ontario RTA
[13] I-Hung Hsu, Zifeng Wang, Long T. Le, Lesly Miculicich, Nanyun Peng, Chen-Yu Lee, and Tomas Pfister. CaLM: Contrasting large and small language models to verify grounded generation. 2024. URL https://arxiv.org/abs/2406.05365. [14] Valentin Noël, Elimane Yassine Seidou, Charly Ken Capo-Chichi, and Ghanem Amari. HalluGraph: Auditable hallucination detection for legal RAG systems via knowledge graph alignment. 2025. URL https://arxiv.org/abs/2512.01659. [15] Piotr Komorowski, Elena Golimblevskaia, Reduan Achtibat, Thomas Wiegand, Sebastian Lapuschkin, and Wojciech Samek. Attribution-Guided Decoding. 2025. URL https://arxiv. org/abs/2509.26307. [16] Shubham Kumar Nigam, Balaramamahanthi Deepak Patnaik, Shivam Mishra, Ajay Varghese Thomas, Noel Shallum, Kripabandhu Ghosh, and Arnab Bhattacharya. NyayaRAG: Realistic legal judgment prediction with RAG under the indian common law system. 2025. URL https://arxiv.org/abs/2508.00709. [17] Darshita Rathore, Vineet Kumar, Chetna Bansal, and Anindya Moitra. How Much is Too Much? exploring LoRA rank trade-offs for retaining knowledge and domain robustness. 2025. URL https://arxiv.org/abs/2512.15634.
10