Adaptive Decoding via Test-Time Policy Learning for Self-Improving Generation

• URL: http://arxiv.org/abs/2603.18428v1
• Date: Thu, 19 Mar 2026 02:44:42 GMT
• Title: Adaptive Decoding via Test-Time Policy Learning for Self-Improving Generation
• Abstract summary: We introduce a reinforcement learning-based sampler that treats decoding as sequential decision-making and learns a lightweight policy to adjust sampling parameters at test-time.<n>We evaluate summarization datasets including BookSum, arXiv, and WikiHow.
• Score: 0.5199807441687141
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Decoding strategies largely determine the quality of Large Language Model (LLM) outputs, yet widely used heuristics such as greedy or fixed temperature/top-p decoding are static and often task-agnostic, leading to suboptimal or inconsistent generation quality across domains that demand stylistic or structural flexibility. We introduce a reinforcement learning-based decoder sampler that treats decoding as sequential decision-making and learns a lightweight policy to adjust sampling parameters at test-time while keeping LLM weights frozen. We evaluated summarization datasets including BookSum, arXiv, and WikiHow using Granite-3.3-2B and Qwen-2.5-0.5B. Our policy sampler consistently outperforms greedy and static baselines, achieving relative gains of up to +88% (BookSum, Granite) and +79% (WikiHow, Qwen). Reward ablations show that overlap-only objectives underperform compared to composite rewards, while structured shaping terms (length, coverage, repetition, completeness) enable stable and sustained improvements. These findings highlight reinforcement learning as a practical mechanism for test-time adaptation in decoding, enabling domain-aware and user-controllable generation without retraining large models.

Related papers

• Unsupervised Continual Learning with Growing Self-Organizing Maps and Synthetic Replay [47.020758735910306]
  This work presents a generative continual learning framework based on growing self-organizing maps (GSOMs) and encoder-decoder models for class-incremental learning.<n>The proposed approach enables exemplar-free replay using distributional statistical memory, which eliminates the need to store raw data.<n>Results across multiple benchmarks show that the proposed approach achieves performance competitive even with supervised state-of-the-art memory-based methods.
  arXiv  Detail & Related papers  (2026-08-27T19:48:44Z)
• Multimodal Learning on Low-Quality Data with Conformal Predictive Self-Calibration [72.0672328514289]
  Multimodal learning often grapples with the challenge of low-quality data.<n>We propose a unified framework, termed Conformal Predictive Self-Calibration.<n>Our framework consistently outperforms existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-05-05T14:48:52Z)
• Baguan-TS: A Sequence-Native In-Context Learning Model for Time Series Forecasting with Covariates [31.296823831987748]
  Baguan-TS integrates the raw-sequence representation learning with ICL, instantiated by a 3D Transformer.<n>We tackle two key hurdles: (i) calibration and training stability, improved with a feature-agnostic, target-space retrieval-based local calibration, and (ii) output oversmoothing, mitigated via context-overfitting strategy.
  arXiv  Detail & Related papers  (2026-03-18T07:24:19Z)
• VADE: Variance-Aware Dynamic Sampling via Online Sample-Level Difficulty Estimation for Multimodal RL [38.782188833641676]
  Group-based policy optimization methods like GRPO and GSPO have become standard for training multimodal models.<n>They suffer from a critical emphgradient vanishing problem when all responses within a group receive identical rewards.<n>We propose textbfVADE, a sampling framework via online sample-level difficulty textbfEstimation.
  arXiv  Detail & Related papers  (2025-11-24T08:59:54Z)
• PointMAC: Meta-Learned Adaptation for Robust Test-Time Point Cloud Completion [16.852116353523257]
  Point cloud completion is essential for robust 3D perception in safety-critical applications such as robotics and augmented reality.<n>Existing models perform static inference and rely heavily on inductive biases learned during training.<n>We propose PointMAC, a meta-learned framework for robust test-time adaptation in point cloud completion.
  arXiv  Detail & Related papers  (2025-10-11T23:13:17Z)
• Unsupervised Online 3D Instance Segmentation with Synthetic Sequences and Dynamic Loss [52.28880405119483]
  Unsupervised online 3D instance segmentation is a fundamental yet challenging task.<n>Existing methods, such as UNIT, have made progress in this direction but remain constrained by limited training diversity.<n>We propose a new framework that enriches the training distribution through synthetic point cloud sequence generation.
  arXiv  Detail & Related papers  (2025-09-27T08:53:27Z)
• CAAD: Context-Aware Adaptive Decoding for Truthful Text Generation [31.469511576774252]
  We propose a context-aware adaptive decoding method for large language models.<n>Our approach achieves a 2.8 percent average improvement on TruthfulQA.<n>Our model-agnostic, scalable, and efficient method requires only a single generation pass.
  arXiv  Detail & Related papers  (2025-08-04T08:28:25Z)
• Inference-Time Scaling of Diffusion Language Models with Particle Gibbs Sampling [70.8832906871441]
  We study how to steer generation toward desired rewards without retraining the models.<n>Prior methods typically resample or filter within a single denoising trajectory, optimizing rewards step-by-step without trajectory-level refinement.<n>We introduce particle Gibbs sampling for diffusion language models (PG-DLM), a novel inference-time algorithm enabling trajectory-level refinement while preserving generation perplexity.
  arXiv  Detail & Related papers  (2025-07-11T08:00:47Z)
• Uncertainty Aware Learning for Language Model Alignment [97.36361196793929]
  We propose uncertainty-aware learning (UAL) to improve the model alignment of different task scenarios. We implement UAL in a simple fashion -- adaptively setting the label smoothing value of training according to the uncertainty of individual samples. Experiments on widely used benchmarks demonstrate that our UAL significantly and consistently outperforms standard supervised fine-tuning.
  arXiv  Detail & Related papers  (2024-06-07T11:37:45Z)
• Towards Continual Learning Desiderata via HSIC-Bottleneck Orthogonalization and Equiangular Embedding [55.107555305760954]
  We propose a conceptually simple yet effective method that attributes forgetting to layer-wise parameter overwriting and the resulting decision boundary distortion. Our method achieves competitive accuracy performance, even with absolute superiority of zero exemplar buffer and 1.02x the base model.
  arXiv  Detail & Related papers  (2024-01-17T09:01:29Z)
• Latent Template Induction with Gumbel-CRFs [107.17408593510372]
  We explore the use of structured variational autoencoders to infer latent templates for sentence generation. As a structured inference network, we show that it learns interpretable templates during training.
  arXiv  Detail & Related papers  (2020-11-29T01:00:57Z)
• AdaS: Adaptive Scheduling of Stochastic Gradients [50.80697760166045]
  We introduce the notions of textit"knowledge gain" and textit"mapping condition" and propose a new algorithm called Adaptive Scheduling (AdaS) Experimentation reveals that, using the derived metrics, AdaS exhibits: (a) faster convergence and superior generalization over existing adaptive learning methods; and (b) lack of dependence on a validation set to determine when to stop training.
  arXiv  Detail & Related papers  (2020-06-11T16:36:31Z)

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.