Do Small Models Use the Law You Give Them? Measuring Context Use on a Bilingual Bangladesh Legal Benchmark

• URL: http://arxiv.org/abs/2608.30327v1
• Date: Mon, 31 Aug 2026 06:45:16 GMT
• Title: Do Small Models Use the Law You Give Them? Measuring Context Use on a Bilingual Bangladesh Legal Benchmark
• Abstract summary: Fine-tuning can improve legal question-answering accuracy without improving how models use law supplied in context.<n>We construct a hierarchy-preserving statutory corpus, 2,165 reviewed bilingual fine-tuning examples, and a 150-item supplied-law control.
• Score: 2.062723244880136
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Fine-tuning can improve legal question-answering accuracy without improving how models use law supplied in context. We study this distinction in bilingual Bangladeshi legal QA, where observed errors can arise from answer scoring, retrieval, or failure to use relevant law. We construct a hierarchy-preserving statutory corpus, 2,165 reviewed bilingual fine-tuning examples, and a 150-item supplied-law control. We evaluate six instruction-tuned models: Llama-3.2-1B, Llama-3.2-3B, Qwen3.5-0.8B, Qwen3.5-2B, Qwen3.5-4B, and Gemma-4-E2B, with three LoRA seeds per model. To separate effects, we combine constrained option-letter scoring, cyclic option rotation, and controlled removal of the governing provision. On 398 Bar Council outputs, an exact-line parser attributes an accuracy gain of 50.0\% to the Qwen3.5-2B seed-42 adapter, whereas option scoring yields only $3.0\%$. For Gemma-4-E2B, the two scoring methods favor different systems. When the governing provision is guaranteed to be present, five of six reference models improve by $14.7\%-19.3\%$ under the four-order criterion. Removing that provision reduces accuracy by $8.0\%-15.3\%$ for models and by $13.8\%-14.9\%$ points for their adapters. However, difference-in differences estimates show no increase in reliance on the governing provision after fine-tuning. Results show that legal adaptation claims require separating scorer, retriever, and model effects. Our Code and data are available at https://anonymous.4open.science/r/bangladesh-legal-qa-11E3

Related papers

• Do Small Models Use the Law You Give Them? Context-Injected Fine-Tuning for Legal QA in Bangladesh [2.062723244880136]
  A small language model can receive the governing statutory provision and still answer incorrectly.<n>We test whether fine-tuning on examples containing relevant law improves later use of retrieved law.<n> Evaluation uses the 2022 and 2023 Bangladesh Bar Council exams in Bangla and machine-translated English.
  arXiv  Detail & Related papers  (2026-07-26T04:06:13Z)
• Confidently Wrong: Exception Chain Collapse in Frontier LLM Rule Evaluation [0.0]
  We document a failure class in frontier large language models.<n>The failure reproduces at first observation, but its empirical surface is unstable.<n>For regulated, frontier-model accuracy is a moving compliance boundary that shifts without notice.
  arXiv  Detail & Related papers  (2026-07-25T22:40:02Z)
• Abliteration Is Not a Scalpel: Off-Target Effects of Refusal Removal on Decision Disposition Across Model Families [51.56484100374058]
  Abliteration - deleting a model's refusal direction from its weights - is the standard recipe behind popular "uncensored" open-weight models.<n>As a disposition probe we use 21,600 decisions under uncertainty, replayed through a frozen pipeline so the decision-layer model is the only variable.<n>Three effects replicate across both families (weeks-clustered CIs excluding zero)<n>A fourth effect reverses sign: the same operation makes Gemma-abliterated less confident and Qwen-abliterated more (family CIs non-overlapping)
  arXiv  Detail & Related papers  (2026-07-19T22:27:00Z)
• ImmigrationQA: A Source-Grounded Dataset and Small-Model Adaptation for U.S. Immigration Law [0.0]
  We describe the construction of ImmigrationQA, a source-grounded question-answering dataset of 17,058 pairs.<n>The corpus was assembled from 11 primary and secondary sources.<n>The fine-tuned model was evaluated against a held-out split of 993 pairs.
  arXiv  Detail & Related papers  (2026-05-28T21:36:18Z)
• Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment [0.0]
  We evaluate nine open-weight language models between 135M and 3B parameters on a benchmark designed for structured local deployment.<n>A shared parameter-efficient fine-tuning pipeline adapts a subset of models using 4-bit NF4 quantization with DoRA/LoRA-style adapters on an NVIDIA L4-class budget.<n>In base evaluation, Qwen Coder 3B leads at 75.67% strict accuracy, followed by Qwen2.5 1.5B at 67.10%, Qwen3.5 2B at 64.98%, and Granite 3.3 2B at 64.61%.
  arXiv  Detail & Related papers  (2026-05-05T02:36:09Z)
• LegalBench-BR: A Benchmark for Evaluating Large Language Models on Brazilian Legal Decision Classification [0.0]
  We introduce LegalBench-BR, the first public benchmark for evaluating language models on Brazilian legal text classification.<n>The dataset comprises 3,105 appellate proceedings from the Santa Catarina State Court (TJSC)<n>On a class-balanced test set, BERTimbau-LoRA achieves 87.6% accuracy and 0.87 macro-F1 (+22pp over Claude 3.5 Haiku, +28pp over GPT-4o mini)
  arXiv  Detail & Related papers  (2026-04-20T22:00:02Z)
• Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers [90.50039419576807]
  Reinforcement Learning with Verifiable Rewards (RLVR) trains policies against automated verifiers to avoid costly human labeling.<n>To reduce vulnerability to verifier hacking, many RLVR systems collapse rewards to binary $0,1$ during training.<n>This choice carries a cost: it introduces textitfalse negatives (rejecting correct answers, FNs) and textitfalse positives (accepting incorrect ones, FPs)
  arXiv  Detail & Related papers  (2025-10-01T13:56:44Z)
• Variational Low-Rank Adaptation Using IVON [50.0349688602232]
  We show that variational learning can significantly improve the accuracy and calibration of Low-Rank Adaptation (LoRA) We replace AdamW by the Improved Variational Online Newton (IVON) algorithm to finetune large language models.
  arXiv  Detail & Related papers  (2024-11-07T04:17:30Z)
• Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models [61.467781476005435]
  skill-wise performance is obscured when inspecting aggregate accuracy, under-utilizing the rich signal modern benchmarks contain. We propose an automatic approach to recover the underlying skills relevant for any evaluation instance, by way of inspecting model-generated rationales. Our skill-slices and framework open a new avenue in model evaluation, leveraging skill-specific analyses to unlock a more granular and actionable understanding of model capabilities.
  arXiv  Detail & Related papers  (2024-10-17T17:51:40Z)
• Uncertainty in Language Models: Assessment through Rank-Calibration [65.10149293133846]
  Language Models (LMs) have shown promising performance in natural language generation. It is crucial to correctly quantify their uncertainty in responding to given inputs. We develop a novel and practical framework, termed $Rank$-$Calibration$, to assess uncertainty and confidence measures for LMs.
  arXiv  Detail & Related papers  (2024-04-04T02:31:05Z)
• Model Editing with Canonical Examples [75.33218320106585]
  We introduce model editing with canonical examples. A canonical example is a simple instance of good behavior, e.g., The capital of Mauritius is Port Louis. We propose sense finetuning, which selects and finetunes a few sense vectors for each canonical example.
  arXiv  Detail & Related papers  (2024-02-09T03:08:12Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.