Decoupling Task-Solving and Output Formatting in LLM Generation

• URL: http://arxiv.org/abs/2510.03595v1
• Date: Sat, 04 Oct 2025 00:52:48 GMT
• Title: Decoupling Task-Solving and Output Formatting in LLM Generation
• Abstract summary: We introduce Deco-G, a decoding framework that explicitly decouples format adherence from task solving.<n> Deco-G handles format compliance with a separate tractable probabilistic model (TPM)<n>We demonstrate the effectiveness of Deco-G across a wide range of tasks with diverse format requirements.
• Score: 44.40087140333511
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) are increasingly adept at following instructions containing task descriptions to solve complex problems, such as mathematical reasoning and automatic evaluation (LLM-as-a-Judge). However, as prompts grow more complex, models often struggle to adhere to all instructions. This difficulty is especially common when instructive prompts intertwine reasoning directives -- specifying what the model should solve -- with rigid formatting requirements that dictate how the solution must be presented. The entanglement creates competing goals for the model, suggesting that more explicit separation of these two aspects could lead to improved performance. To this front, we introduce Deco-G, a decoding framework that explicitly decouples format adherence from task solving. Deco-G handles format compliance with a separate tractable probabilistic model (TPM), while prompts LLMs with only task instructions. At each decoding step, Deco-G combines next token probabilities from the LLM with the TPM calculated format compliance likelihood to form the output probability. To make this approach both practical and scalable for modern instruction-tuned LLMs, we introduce three key innovations: instruction-aware distillation, a flexible trie-building algorithm, and HMM state pruning for computational efficiency. We demonstrate the effectiveness of Deco-G across a wide range of tasks with diverse format requirements, including mathematical reasoning, LLM-as-a-judge, and event argument extraction. Overall, our approach yields 1.0% to 6.0% relative gain over regular prompting practice with guaranteed format compliance.

Related papers

• Fine-tuning MLLMs Without Forgetting Is Easier Than You Think [72.59321247529975]
  We design a 2x2 experimental framework to assess model performance across in-distribution and out-of-distribution image and text inputs.<n>Our results show that appropriate regularization, such as constraining the number of trainable parameters or adopting a low learning rate, effectively prevents forgetting when dealing with out-of-distribution images.<n>We attribute this forgetting as task-specific overfitting and address this issue by introducing a data-hybrid training strategy.
  arXiv  Detail & Related papers  (2026-03-15T17:16:19Z)
• Mixture of Reasonings: Teach Large Language Models to Reason with Adaptive Strategies [6.7519234849348075]
  Mixture of Reasoning embeds diverse reasoning strategies into large language models.<n>MoR significantly enhances performance, with MoR150 achieving 0.730 (2.2% improvement) using CoT prompting and 0.734 (13.5% improvement) compared to baselines.
  arXiv  Detail & Related papers  (2025-07-01T09:39:04Z)
• SPARE: Single-Pass Annotation with Reference-Guided Evaluation for Automatic Process Supervision and Reward Modelling [58.05959902776133]
  We introduce Single-Pass.<n>with Reference-Guided Evaluation (SPARE), a novel structured framework that enables efficient per-step annotation.<n>We demonstrate SPARE's effectiveness across four diverse datasets spanning mathematical reasoning (GSM8K, MATH), multi-hop question answering (MuSiQue-Ans), and spatial reasoning (SpaRP)<n>On ProcessBench, SPARE demonstrates data-efficient out-of-distribution generalization, using only $sim$16% of training samples compared to human-labeled and other synthetically trained baselines.
  arXiv  Detail & Related papers  (2025-06-18T14:37:59Z)
• Self-Steering Language Models [113.96916935955842]
  DisCIPL is a method for "self-steering" language models (LMs)<n>DisCIPL generates a task-specific inference program that is executed by a population of Follower models.<n>Our work opens up a design space of highly-parallelized Monte Carlo inference strategies.
  arXiv  Detail & Related papers  (2025-04-09T17:54:22Z)
• Improving Consistency in Large Language Models through Chain of Guidance [9.040736633675136]
  Chain of Guidance (CoG) is a multistep prompting technique that generates highly consistent outputs from Large Language Models (LLMs)<n>We use synthetic data sets comprised of consistent input-output pairs to fine-tune LLMs to produce consistent and correct outputs.<n>Our fine-tuned models are more than twice as consistent compared to base models and show strong generalization capabilities by producing consistent outputs over datasets not used in the fine-tuning process.
  arXiv  Detail & Related papers  (2025-02-21T20:41:37Z)
• Interactive and Expressive Code-Augmented Planning with Large Language Models [62.799579304821826]
  Large Language Models (LLMs) demonstrate strong abilities in common-sense reasoning and interactive decision-making. Recent techniques have sought to structure LLM outputs using control flow and other code-adjacent techniques to improve planning performance. We propose REPL-Plan, an LLM planning approach that is fully code-expressive and dynamic.
  arXiv  Detail & Related papers  (2024-11-21T04:23:17Z)
• Gap-Filling Prompting Enhances Code-Assisted Mathematical Reasoning [0.0]
  Chain-of-thought (CoT) and program-of-thought (PoT) fine-tuning are common methods to transfer LLM knowledge to small language models (SLMs) This paper introduces Gap-Filling Prompting (GFP), a novel two-step prompting strategy designed to enhance the problem-solving process for SLMs.
  arXiv  Detail & Related papers  (2024-11-08T08:52:59Z)
• Graph-DPEP: Decomposed Plug and Ensemble Play for Few-Shot Document Relation Extraction with Graph-of-Thoughts Reasoning [34.85741925091139]
  Graph-DPEP framework is grounded in the reasoning behind triplet explanation thoughts presented in natural language. We develop "ensemble-play", reapplying generation on the entire type list by leveraging the reasoning thoughts embedded in a sub-graph.
  arXiv  Detail & Related papers  (2024-11-05T07:12:36Z)
• Can Large Language Models Understand Real-World Complex Instructions? [54.86632921036983]
  Large language models (LLMs) can understand human instructions, but struggle with complex instructions. Existing benchmarks are insufficient to assess LLMs' ability to understand complex instructions. We propose CELLO, a benchmark for evaluating LLMs' ability to follow complex instructions systematically.
  arXiv  Detail & Related papers  (2023-09-17T04:18:39Z)
• SatLM: Satisfiability-Aided Language Models Using Declarative Prompting [68.40726892904286]
  We propose a new satisfiability-aided language modeling (SatLM) approach for improving the reasoning capabilities of large language models (LLMs) We use an LLM to generate a declarative task specification rather than an imperative program and leverage an off-the-shelf automated theorem prover to derive the final answer. We evaluate SATLM on 8 different datasets and show that it consistently outperforms program-aided LMs in the imperative paradigm.
  arXiv  Detail & Related papers  (2023-05-16T17:55:51Z)

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.