Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2604.09737 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908954006650880 |
|---|---|
| author | Fodeh, Samah Puthiaraju, Ganesh Irankhah, Elyas Ma, Linhai Talakokkul, Srivani Khan, Afshan Ramachandran, Sreeraj Alpert, Jordan Schellhorn, Sarah |
| author_facet | Fodeh, Samah Puthiaraju, Ganesh Irankhah, Elyas Ma, Linhai Talakokkul, Srivani Khan, Afshan Ramachandran, Sreeraj Alpert, Jordan Schellhorn, Sarah |
| contents | Structured prediction requires models to generate ontology-constrained labels, grounded evidence, and valid structure under ambiguity, label skew, and heterogeneous group difficulty. We present a two-part framework for controllable inference and robust fine-tuning. First, we introduce a task-agnostic prompting strategy that combines XML-based instruction structure, disambiguation rules, verification-style reasoning, schema constraints, and self-validation to address format drift, label ambiguity, evidence hallucination, and metadata-conditioned confusion in in-context structured generation. Second, we introduce STaR-DRO, a stateful robust optimization method for group heterogeneity. It combines Tsallis mirror descent with momentum-smoothed, centered group-loss signals and bounded excess-only multipliers so that only persistently hard groups above a neutral baseline are upweighted, concentrating learning where it is most needed while avoiding volatile, dense exponentiated-gradient reweighting and unnecessary loss from downweighting easier groups. We evaluate the combined framework on EPPC Miner, a benchmark for extracting hierarchical labels and evidence spans from patient-provider secure messages. Prompt engineering improves zero-shot by +15.44 average F1 across Code, Sub-code, and Span over four Llama models. Building on supervised fine-tuning, STaR-DRO further improves the hardest semantic decisions: on Llama-3.3-70B-Instruct, Code F1 rises from 79.24 to 81.47 and Sub-code F1 from 67.78 to 69.30, while preserving Span performance and reducing group-wise validation cross-entropy by up to 29.6% on the most difficult clinical categories. Because these rare and difficult groups correspond to clinically consequential communication behaviors, these gains are not merely statistical improvements: they directly strengthen communication mining reliability for patient-centered care analysis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_09737 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | STaR-DRO: Stateful Tsallis Reweighting for Group-Robust Structured Prediction Fodeh, Samah Puthiaraju, Ganesh Irankhah, Elyas Ma, Linhai Talakokkul, Srivani Khan, Afshan Ramachandran, Sreeraj Alpert, Jordan Schellhorn, Sarah Machine Learning Artificial Intelligence Structured prediction requires models to generate ontology-constrained labels, grounded evidence, and valid structure under ambiguity, label skew, and heterogeneous group difficulty. We present a two-part framework for controllable inference and robust fine-tuning. First, we introduce a task-agnostic prompting strategy that combines XML-based instruction structure, disambiguation rules, verification-style reasoning, schema constraints, and self-validation to address format drift, label ambiguity, evidence hallucination, and metadata-conditioned confusion in in-context structured generation. Second, we introduce STaR-DRO, a stateful robust optimization method for group heterogeneity. It combines Tsallis mirror descent with momentum-smoothed, centered group-loss signals and bounded excess-only multipliers so that only persistently hard groups above a neutral baseline are upweighted, concentrating learning where it is most needed while avoiding volatile, dense exponentiated-gradient reweighting and unnecessary loss from downweighting easier groups. We evaluate the combined framework on EPPC Miner, a benchmark for extracting hierarchical labels and evidence spans from patient-provider secure messages. Prompt engineering improves zero-shot by +15.44 average F1 across Code, Sub-code, and Span over four Llama models. Building on supervised fine-tuning, STaR-DRO further improves the hardest semantic decisions: on Llama-3.3-70B-Instruct, Code F1 rises from 79.24 to 81.47 and Sub-code F1 from 67.78 to 69.30, while preserving Span performance and reducing group-wise validation cross-entropy by up to 29.6% on the most difficult clinical categories. Because these rare and difficult groups correspond to clinically consequential communication behaviors, these gains are not merely statistical improvements: they directly strengthen communication mining reliability for patient-centered care analysis. |
| title | STaR-DRO: Stateful Tsallis Reweighting for Group-Robust Structured Prediction |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2604.09737 |