$φ$-Decoding: Adaptive Foresight Sampling for Balanced Inference-Time Exploration and Exploitation

• URL: http://arxiv.org/abs/2503.13288v1
• Date: Mon, 17 Mar 2025 15:38:33 GMT
• Title: $φ$-Decoding: Adaptive Foresight Sampling for Balanced Inference-Time Exploration and Exploitation
• Authors: Fangzhi Xu, Hang Yan, Chang Ma, Haiteng Zhao, Jun Liu, Qika Lin, Zhiyong Wu,
• Abstract summary: In-time optimization scales computation to derive deliberate reasoning steps for effective performance.<n>We frame the decoding strategy as foresight sampling, leveraging simulated future steps to obtain globally optimal step estimation.<n>Experiments show $phi$-Decoding outperforms strong baselines in both performance and efficiency.
• Score: 22.607133083903125
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Inference-time optimization scales computation to derive deliberate reasoning steps for effective performance. While previous search-based strategies address the short-sightedness of auto-regressive generation, the vast search space leads to excessive exploration and insufficient exploitation. To strike an efficient balance to derive the optimal step, we frame the decoding strategy as foresight sampling, leveraging simulated future steps to obtain globally optimal step estimation. Built on it, we propose a novel decoding strategy, named $\phi$-Decoding. To provide a precise and expressive estimation of step value, $\phi$-Decoding approximates two distributions via foresight and clustering. Sampling from the joint distribution, the optimal steps can be selected for exploitation. To support adaptive computation allocation, we propose in-width and in-depth pruning strategies, featuring a light-weight solution to achieve inference efficiency. Extensive experiments across seven benchmarks show $\phi$-Decoding outperforms strong baselines in both performance and efficiency. Additional analysis demonstrates its generalization across various LLMs and scalability across a wide range of computing budgets. The code will be released at https://github.com/xufangzhi/phi-Decoding, and the open-source PyPI package is coming soon.

Related papers

• Review, Refine, Repeat: Understanding Iterative Decoding of AI Agents with Dynamic Evaluation and Selection [71.92083784393418]
  Inference-time methods such as Best-of-N (BON) sampling offer a simple yet effective alternative to improve performance. We propose Iterative Agent Decoding (IAD) which combines iterative refinement with dynamic candidate evaluation and selection guided by a verifier.
  arXiv  Detail & Related papers  (2025-04-02T17:40:47Z)
• DARS: Dynamic Action Re-Sampling to Enhance Coding Agent Performance by Adaptive Tree Traversal [55.13854171147104]
  Large Language Models (LLMs) have revolutionized various domains, including natural language processing, data analysis, and software development. We present Dynamic Action Re-Sampling (DARS), a novel inference time compute scaling approach for coding agents. We evaluate our approach on SWE-Bench Lite benchmark, demonstrating that this scaling strategy achieves a pass@k score of 55% with Claude 3.5 Sonnet V2.
  arXiv  Detail & Related papers  (2025-03-18T14:02:59Z)
• Optimistic ε-Greedy Exploration for Cooperative Multi-Agent Reinforcement Learning [16.049852176246038]
  We propose Optimistic $epsilon$-Greedy Exploration, focusing on enhancing exploration to correct value estimations.<n>We introduce an optimistic updating network to identify optimal actions and sample actions from its distribution with a probability of $epsilon$ during exploration.<n> Experimental results in various environments reveal that the Optimistic $epsilon$-Greedy Exploration effectively prevents the algorithm from suboptimal solutions.
  arXiv  Detail & Related papers  (2025-02-05T12:06:54Z)
• Efficient Learning of POMDPs with Known Observation Model in Average-Reward Setting [56.92178753201331]
  We propose the Observation-Aware Spectral (OAS) estimation technique, which enables the POMDP parameters to be learned from samples collected using a belief-based policy. We show the consistency of the OAS procedure, and we prove a regret guarantee of order $mathcalO(sqrtT log(T)$ for the proposed OAS-UCRL algorithm.
  arXiv  Detail & Related papers  (2024-10-02T08:46:34Z)
• Discovering Preference Optimization Algorithms with and for Large Language Models [50.843710797024805]
  offline preference optimization is a key method for enhancing and controlling the quality of Large Language Model (LLM) outputs. We perform objective discovery to automatically discover new state-of-the-art preference optimization algorithms without (expert) human intervention. Experiments demonstrate the state-of-the-art performance of DiscoPOP, a novel algorithm that adaptively blends logistic and exponential losses.
  arXiv  Detail & Related papers  (2024-06-12T16:58:41Z)
• A Scalable and Near-Optimal Conformance Checking Approach for Long Traces [3.3170150440851485]
  Conformity checking, a key task in process mining, can become computationally infeasible due to the exponential complexity of finding an optimal alignment. This paper introduces a novel sliding window approach to address these scalability challenges. By breaking down traces into manageable subtraces and iteratively aligning each with the process model, our method significantly reduces the search space.
  arXiv  Detail & Related papers  (2024-06-08T11:04:42Z)
• Beyond Single-Model Views for Deep Learning: Optimization versus Generalizability of Stochastic Optimization Algorithms [13.134564730161983]
  This paper adopts a novel approach to deep learning optimization, focusing on gradient descent (SGD) and its variants. We show that SGD and its variants demonstrate performance on par with flat-minimas like SAM, albeit with half the gradient evaluations. Our study uncovers several key findings regarding the relationship between training loss and hold-out accuracy, as well as the comparable performance of SGD and noise-enabled variants.
  arXiv  Detail & Related papers  (2024-03-01T14:55:22Z)
• Maximize to Explore: One Objective Function Fusing Estimation, Planning, and Exploration [87.53543137162488]
  We propose an easy-to-implement online reinforcement learning (online RL) framework called textttMEX. textttMEX integrates estimation and planning components while balancing exploration exploitation automatically. It can outperform baselines by a stable margin in various MuJoCo environments with sparse rewards.
  arXiv  Detail & Related papers  (2023-05-29T17:25:26Z)
• Truncating Trajectories in Monte Carlo Reinforcement Learning [48.97155920826079]
  In Reinforcement Learning (RL), an agent acts in an unknown environment to maximize the expected cumulative discounted sum of an external reward signal. We propose an a-priori budget allocation strategy that leads to the collection of trajectories of different lengths. We show that an appropriate truncation of the trajectories can succeed in improving performance.
  arXiv  Detail & Related papers  (2023-05-07T19:41:57Z)
• Resource Aware Multifidelity Active Learning for Efficient Optimization [0.8717253904965373]
  This paper introduces the Resource Aware Active Learning (RAAL) strategy to accelerate the optimization of black box functions. The RAAL strategy optimally seeds multiple points at each allowing for a major speed up of the optimization task.
  arXiv  Detail & Related papers  (2020-07-09T10:01:32Z)

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.