paper-with-me

Papers

TreeBoN: Enhancing Inference-Time Alignment with Speculative Tree-Search and Best-of-N Sampling

2024-10-18 · Jiahao Qiu, Yifu Lu, Yifan Zeng, Jiacheng Guo, Jiayi Geng, Huazheng Wang, Kaixuan Huang, Yue Wu, Mengdi Wang

Inference-time alignment enhances the performance of large language models without requiring additional training or fine-tuning but presents challenges due to balancing computational efficiency with high-quality output. Best-of-N (BoN) sampling, as a simple yet powerful approach, generates multiple responses and selects the best one, achieving improved performance but with a high computational cost. We propose TreeBoN, a novel framework that integrates a speculative tree-search strategy into Best-of-N (BoN) Sampling. TreeBoN maintains a set of parent nodes, iteratively branching and pruning low-quality responses, thereby reducing computational overhead while maintaining high output quality. Our approach also leverages token-level rewards from Direct Preference Optimization (DPO) to guide tree expansion and prune low-quality paths. We evaluate TreeBoN using AlpacaFarm, HH-RLHF, UltraFeedback, GSM8K, and TutorEval datasets, demonstrating consistent improvements. Specifically, TreeBoN achieves the highest win rate of 65% on TutorEval and around 60% win rates across other different datasets, outperforming standard BoN with the same computational cost and showcasing its scalability and alignment efficacy.

📄 PDF Abstract BibTeX arXiv:2410.16033

Code (0)

등록된 구현이 없습니다.

Tasks

Computational EfficiencyGSM8K

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically
Pruning 설명 없음

Similar Papers 제목 키워드 기반

Reward-Shifted Speculative Sampling Is An Efficient Test-Time Weak-to-Strong Aligner

2025-08-20 · Bolian Li, Yanran Wu, Xinyu Luo, Ruqi Zhang arxiv

Aligning large language models (LLMs) with human preferences has become a critical step in their development. Recent research has increasingly focused on test-time alignment, where additional compute is allocated during …

Fast Best-of-N Decoding via Speculative Rejection

2024-10-26 · Hanshi Sun, Momin Haider, Ruiqi Zhang, Huitao Yang 외

The safe and effective deployment of Large Language Models (LLMs) involves a critical step called alignment, which ensures that the model's responses are in accordance with human preferences. Prevalent alignment techniqu…

CORAL: Learning Consistent Representations across Multi-step Training with Lighter Speculative Drafter

2025-02-24 · Yepeng Weng, Dianwen Mei, Huishi Qiu, Xujie Chen 외

Speculative decoding is a powerful technique that accelerates Large Language Model (LLM) inference by leveraging a lightweight speculative draft model. However, existing designs suffers in performance due to misalignment…

Large Language Model

Guided Speculative Inference for Efficient Test-Time Alignment of LLMs

2025-06-04 · Jonathan Geuter, Youssef Mroueh, David Alvarez-Melis

We propose Guided Speculative Inference (GSI), a novel algorithm for efficient reward-guided decoding in large language models. GSI combines soft best-of-$n$ test-time scaling with a reward model $r(x,y)$ and speculative…

Math

M2R2: Mixture of Multi-Rate Residuals for Efficient Transformer Inference

2025-02-04 · Nikhil Bhendawade, Mahyar Najibi, Devang Naik, Irina Belousova

Residual transformations enhance the representational depth and expressive power of large language models (LLMs). However, applying static residual transformations across all tokens in auto-regressive generation leads to…

Mixture-of-Experts