Paper Detail

Thought-Level Beam Search for Reasoning

Lijie Yang, Hongyin Luo, Jiawei Zhao, Tri Dao, Ravi Netravali

huggingface Score 15.0

Published 2026-08-11 · First seen 2026-08-15

General AI

Abstract

Test-time compute scaling is a primary driver of performance in large reasoning models (LRMs), but extreme inefficiency bounds current approaches, shifting the critical question from how much compute to spend, to where to allocate it. We formalize test-time reasoning as a constrained compute allocation problem over partial trajectories. Under a fixed hardware budget, existing paradigms fail to actively allocate the compute to the most promising partial progress: traditional parallel sampling treats traces independently and induces severe memory bottlenecks, while subtractive pruning starves hardware and fails to actively and sufficiently shift the output distribution. To overcome this dichotomy, we introduce Gambit, an inference algorithm that executes thought-level beam search. By periodically pruning unpromising trajectories and immediately branching from high-quality prefixes, Gambit dynamically concentrates compute onto the most promising reasoning traces via a light-weight scorer probing hidden states while maintaining continuous high hardware utilization. Extensive evaluations across multiple models and benchmarks demonstrate that Gambit strictly dominates existing baselines. Under identical hardware constraints, our method yields up to a +6.7\% absolute accuracy gain on HMMT-24 and +3.3\% on AIME-25 over pruning baselines, delivers >2times higher throughput on trace completion, and reduces total token consumption by up to 68.5\% relative to standard parallel sampling.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@misc{yang2026thought,
  title = {Thought-Level Beam Search for Reasoning},
  author = {Lijie Yang and Hongyin Luo and Jiawei Zhao and Tri Dao and Ravi Netravali},
  year = {2026},
  abstract = {Test-time compute scaling is a primary driver of performance in large reasoning models (LRMs), but extreme inefficiency bounds current approaches, shifting the critical question from how much compute to spend, to where to allocate it. We formalize test-time reasoning as a constrained compute allocation problem over partial trajectories. Under a fixed hardware budget, existing paradigms fail to actively allocate the compute to the most promising partial progress: traditional parallel sampling tre},
  url = {https://huggingface.co/papers/2608.08020},
  keywords = {test-time compute scaling, large reasoning models, constrained compute allocation, partial trajectories, parallel sampling, subtractive pruning, thought-level beam search, Gambit, hidden states, hardware utilization, code available, huggingface daily},
  eprint = {2608.08020},
  archiveprefix = {arXiv},
}

Metadata

{}