FinVector-Market-4B: A Controlled Study of LoRA Adaptation for Structured Financial Tasks

Key points

• We evaluate the base and adapted models on the same 600-example benchmark under implicit and explicit JSON-schema contracts.1
• Under matched explicit prompting, the frozen scores improve from 14.7% to 40.0% for FinQA answer exact match, from 48.0% to 82.7% for calculator-expression correctness, from 20.1% to 89.5% for scenario branch-label agreement, and from 52.4% to 87.2% for implication-direction agreement.1
• Filing overlap and calculator-target inconsistencies qualify the benchmark's generalization claims.1
• The results show that compact financial domain adaptation can produce substantial task-specific gains beyond output-format learning under matched prompting, with gains bounded by the evaluated task distribution and prompt contract.1

Sources (1)

• [1]FinVector-Market-4B: A Controlled Study of LoRA Adaptation for Structured Financial Tasks
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 10:46 AM

  FinVector-Market-4B adapts Qwen/Qwen3.5-4B with rank-16 LoRA on a 22,000-example corpus for structured financial tasks.

  We evaluate the base and adapted models on the same 600-example benchmark under implicit and explicit JSON-schema contracts.

Extractive summary: sentences quoted from the sources.