Work

When LLM Meets Tree Search: A Systematic View of Inference as Search in Large Language Models

Year: 2026

Type: conference-paper

Abstract: As pretraining scaling laws approach saturation, Test-Time Scaling (TTS) has emerged as an important direction for improving reasoning by allocating inference-time compute to a fixed model prior.Viewed at a high level, TTS reframes inference as search over a space of partial reasoning states.While Chainof-Thought (CoT) exposes intermediate steps, common instantiations rely on single-trajectory decoding, limiting recovery from early errors and exploration.This survey systematizes recent progress in tree-search-based reasoning, viewing inference as instance-specific optimization rather than decoding.We trace the evolution from uninformed search to Monte Carlo Tree Search (MCTS), highlighting how sampling-based control supports principled exploration-exploitation trade-offs.To unify a fragmented literature, we introduce a Unified Design Space spanning search topology, evaluation signals, and control dynamics, and advocate a standardized compute-reporting abstraction to make compute-accuracy tradeoffs explicit and comparable.

Source: Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)

DOI: 10.18653/2026.findings-emnlp.643

Authors: Jiaqi Wei, Xiang Zhang, Yuejin Yang, Wenxuan Huang, Juntai Cao, Sheng Xu, Xiang Zhuang, Zhangyang Gao, Muhammad Abdul-Mageed, Laks V. S. Lakshmanan, Chenyu You, Wanli Ouyang, Siqi Sun

Institutions: Shanghai Artificial Intelligence Laboratory, Zhejiang University, University of British Columbia, Fudan University, Stony Brook University, Chinese University of Hong Kong

Language: English

Topic: Topic Modeling

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: bronze

OpenAlex ID: W7204907974

FWCI:

Cites:

Cited by:

Related to: 10