Adaptive Test-Time Compute Allocation with Evolving In-Context Demonstrations

• URL: http://arxiv.org/abs/2604.21018v1
• Date: Wed, 22 Apr 2026 19:07:29 GMT
• Title: Adaptive Test-Time Compute Allocation with Evolving In-Context Demonstrations
• Abstract summary: We introduce a test-time compute allocation framework that jointly adapts where computation is spent and how generation is performed.<n>Our approach consistently outperforms existing baselines while consuming substantially less inference-time compute.
• Score: 25.993304809118726
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While scaling test-time compute can substantially improve model performance, existing approaches either rely on static compute allocation or sample from fixed generation distributions. In this work, we introduce a test-time compute allocation framework that jointly adapts where computation is spent and how generation is performed. Our method begins with a warm-up phase that identifies easy queries and assembles an initial pool of question-response pairs from the test set itself. An adaptive phase then concentrates further computation on unresolved queries while reshaping their generation distributions through evolving in-context demonstrations -- conditioning each generation on successful responses from semantically related queries rather than resampling from a fixed distribution. Experiments across math, coding, and reasoning benchmarks demonstrate that our approach consistently outperforms existing baselines while consuming substantially less inference-time compute.

Related papers

• Deciding When to Switch: E-Processes for Adaptive Minimax Training for Generative Adversarial Nets [8.044490027380872]
  We develop an e-process-based adaptive training procedure for adversarial networks.<n>We prove that fresh empirical and latent draws yield conditional e-values that can be accumulated into e-processes.<n>Across synthetic distributions and image benchmark datasets, the proposed method matches or outperforms the best fixed-ratio baselines.
  arXiv  Detail & Related papers  (2026-08-10T18:07:48Z)
• $V_1$: Unifying Generation and Self-Verification for Parallel Reasoners [69.66089681814013]
  $V_$ is a framework that unifies generation and verification through efficient pairwise ranking.<n>$V_$-Infer improves Pass@1 by up to $10%$ over pointwise verification.<n>$V_$-PairRL achieves $7$--$9%$ test-time scaling gains over standard RL and pointwise joint training.
  arXiv  Detail & Related papers  (2026-03-04T17:22:16Z)
• What If We Allocate Test-Time Compute Adaptively? [2.1713977971908944]
  Test-time scaling allocates inference computation uniformly, uses fixed sampling strategies, and applies verification only for reranking.<n>We propose a verifier-guided adaptive framework treating reasoning as iterative trajectory generation and selection.<n>Across datasets, our dynamic, PRM-guided approach consistently outperforms direct test-time scaling.
  arXiv  Detail & Related papers  (2026-02-01T07:30:22Z)
• Latency and Token-Aware Test-Time Compute [3.573250939705335]
  Inference-time scaling can improve large language model (LLM) performance by generating multiple candidate responses and selecting among them.<n>We formulate inference-time scaling as a problem of dynamic compute allocation and method selection.<n>Our framework explicitly incorporates both token cost and wall-clock latency, the latter being critical for user experience and particularly for agentic models.
  arXiv  Detail & Related papers  (2025-09-11T21:35:19Z)
• Sample, Don't Search: Rethinking Test-Time Alignment for Language Models [55.2480439325792]
  We introduce QAlign, a new test-time alignment approach.<n>As we scale test-time compute, QAlign converges to sampling from the optimal aligned distribution for each individual prompt.<n>By adopting recent advances in Markov chain Monte Carlo for text generation, our method enables better-aligned outputs without modifying the underlying model or even requiring logit access.
  arXiv  Detail & Related papers  (2025-04-04T00:41:40Z)
• IT$^3$: Idempotent Test-Time Training [95.78053599609044]
  Deep learning models often struggle when deployed in real-world settings due to distribution shifts between training and test data.<n>We present Idempotent Test-Time Training (IT$3$), a novel approach that enables on-the-fly adaptation to distribution shifts using only the current test instance.<n>Our results suggest that idempotence provides a universal principle for test-time adaptation that generalizes across domains and architectures.
  arXiv  Detail & Related papers  (2024-10-05T15:39:51Z)
• DOTA: Distributional Test-Time Adaptation of Vision-Language Models [69.41389326333771]
  Vision-language foundation models can be unreliable when significant distribution gaps exist between training and test data.<n>We propose DOTA (DistributiOnal Test-time Adaptation), a simple yet effective method addressing this limitation.<n>This distribution-centric approach enables the model to continually learn and adapt to the deployment environment.
  arXiv  Detail & Related papers  (2024-09-28T15:03:28Z)
• Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approach [14.958884168060097]
  We present a novel approach for test-time adaptation via online self-training.<n>Our approach combines concepts in betting martingales and online learning to form a detection tool capable of reacting to distribution shifts.<n> Experimental results demonstrate that our approach improves test-time accuracy under distribution shifts while maintaining accuracy and calibration in their absence.
  arXiv  Detail & Related papers  (2024-08-14T12:40:57Z)
• Adaptive scheduling for adaptive sampling in POS taggers construction [0.27624021966289597]
  We introduce an adaptive scheduling for adaptive sampling as a novel way of machine learning in the construction of part-of-speech taggers. We analyze the shape of the learning curve geometrically in conjunction with a functional model to increase or decrease it at any time. We also improve the robustness of sampling by paying greater attention to those regions of the training data base subject to a temporary inflation in performance.
  arXiv  Detail & Related papers  (2024-02-04T15:02:17Z)
• Noisy Correspondence Learning with Self-Reinforcing Errors Mitigation [63.180725016463974]
  Cross-modal retrieval relies on well-matched large-scale datasets that are laborious in practice. We introduce a novel noisy correspondence learning framework, namely textbfSelf-textbfReinforcing textbfErrors textbfMitigation (SREM)
  arXiv  Detail & Related papers  (2023-12-27T09:03:43Z)
• Time-series Generation by Contrastive Imitation [87.51882102248395]
  We study a generative framework that seeks to combine the strengths of both: Motivated by a moment-matching objective to mitigate compounding error, we optimize a local (but forward-looking) transition policy. At inference, the learned policy serves as the generator for iterative sampling, and the learned energy serves as a trajectory-level measure for evaluating sample quality.
  arXiv  Detail & Related papers  (2023-11-02T16:45:25Z)
• Semantic Self-adaptation: Enhancing Generalization with a Single Sample [45.111358665370524]
  We propose a self-adaptive approach for semantic segmentation. It fine-tunes the parameters of convolutional layers to the input image using consistency regularization. Our empirical study suggests that self-adaptation may complement the established practice of model regularization at training time.
  arXiv  Detail & Related papers  (2022-08-10T12:29:01Z)

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.