Back to AI News
arxiv

Parameter-Efficient Conciseness Adaptation: Differentiable Length Control for Concise Clinical Reasoning in Smaller Language Models

The paper introduces PECA, a method that adds a differentiable conciseness term to QLoRA supervised fine-tuning for smaller language models. Instead of relying on preference data, reward models, or proprietary teachers, PECA raises end-of-sequence probability once generation exceeds a token budget. On a 1.7-billion-parameter model fine-tuned on the medical-o1-reasoning-SFT corpus, PECA reduced mean response length by 72.8% while improving multiple-choice diagnostic accuracy from 40.8% to 44.0%. The authors also describe a sweep over conciseness strength that traces an accuracy-to-length frontier.

AI Brief

It offers a concrete, low-cost training technique for making small clinical models less verbose without sacrificing diagnostic accuracy, which is useful for point-of-care deployment.

arXiv:2611.17128v1 Announce Type: cross Abstract: Smaller language models (SLMs) make it feasible to deploy clinical natural language processing where large proprietary models are too costly, and parameter-efficient fine-tuning allows them to be specialised on commodity hardware. One obstacle persists after training: on medical tasks these models answer at far greater length than the clinical question requires. In clinical decision support and point-of-care use, where time is short, such verbosity is hard to act on even when the underlying reasoning is sound. Existing length-control methods depend on preference optimisation, reward models, or proprietary teachers that undercut the accessibility making SLMs attractive. We introduce Parameter-Efficient Conciseness Adaptation (PECA), which adds a single differentiable conciseness term to the QLoRA supervised fine-tuning objective. Because the token count is discrete and non-differentiable, PECA instead raises the end-of-sequence probability once generation exceeds a token budget, a smooth signal optimised during ordinary training, with no preference data, no auxiliary model, and a single commodity GPU. On a 1.7-billion-parameter model fine-tuned on the medical-o1-reasoning-SFT corpus, PECA reduces mean response length by 72.8% while improving multiple-choice diagnostic accuracy from 40.8% to 44.0%. A sweep over the conciseness strength traces an interpretable accuracy-to-length frontier for balancing brevity and diagnostic completeness.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!