Type-Compliant Adaptation Cascades: Adapting Programmatic LM Workflows to Data

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lin, Chu-Cheng, Peng, Daiyi, Lu, Yifeng, Zhang, Ming, Ie, Eugene
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916972744146944
author Lin, Chu-Cheng
Peng, Daiyi
Lu, Yifeng
Zhang, Ming
Ie, Eugene
author_facet Lin, Chu-Cheng
Peng, Daiyi
Lu, Yifeng
Zhang, Ming
Ie, Eugene
contents Reliably composing Large Language Models (LLMs) for complex, multi-step workflows remains a significant challenge. The dominant paradigm -- optimizing discrete prompts in a pipeline -- is notoriously brittle and struggles to enforce the formal compliance required for structured tasks. We introduce Type-Compliant Adaptation Cascades (TACs), a framework that recasts workflow adaptation as learning typed probabilistic programs. TACs treat the entire workflow, which is composed of parameter-efficiently adapted LLMs and deterministic logic, as an unnormalized joint distribution. This enables principled, gradient-based training even with latent intermediate structures. We provide theoretical justification for our tractable optimization objective, proving that the optimization bias vanishes as the model learns type compliance. Empirically, TACs significantly outperform state-of-the-art prompt-optimization baselines. Gains are particularly pronounced on structured tasks, improving FinQA from $12.0\%$ to $24.7\%$ for a Qwen 3 8B model, MGSM-SymPy from $57.1\%$ to $75.9\%$ for a Gemma 2 27B model, MGSM from $1.6\%$ to $27.3\%$, and MuSR from $36.5\%$ to $62.6\%$ for a Gemma 7B model. TACs offer a robust and theoretically grounded paradigm for developing reliable, task-compliant LLM systems.
format Preprint
id arxiv_https___arxiv_org_abs_2508_18244
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Type-Compliant Adaptation Cascades: Adapting Programmatic LM Workflows to Data
Lin, Chu-Cheng
Peng, Daiyi
Lu, Yifeng
Zhang, Ming
Ie, Eugene
Machine Learning
Artificial Intelligence
Reliably composing Large Language Models (LLMs) for complex, multi-step workflows remains a significant challenge. The dominant paradigm -- optimizing discrete prompts in a pipeline -- is notoriously brittle and struggles to enforce the formal compliance required for structured tasks. We introduce Type-Compliant Adaptation Cascades (TACs), a framework that recasts workflow adaptation as learning typed probabilistic programs. TACs treat the entire workflow, which is composed of parameter-efficiently adapted LLMs and deterministic logic, as an unnormalized joint distribution. This enables principled, gradient-based training even with latent intermediate structures. We provide theoretical justification for our tractable optimization objective, proving that the optimization bias vanishes as the model learns type compliance. Empirically, TACs significantly outperform state-of-the-art prompt-optimization baselines. Gains are particularly pronounced on structured tasks, improving FinQA from $12.0\%$ to $24.7\%$ for a Qwen 3 8B model, MGSM-SymPy from $57.1\%$ to $75.9\%$ for a Gemma 2 27B model, MGSM from $1.6\%$ to $27.3\%$, and MuSR from $36.5\%$ to $62.6\%$ for a Gemma 7B model. TACs offer a robust and theoretically grounded paradigm for developing reliable, task-compliant LLM systems.
title Type-Compliant Adaptation Cascades: Adapting Programmatic LM Workflows to Data
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2508.18244