AI News HubLIVE
原文

PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning

Large reasoning language models (LRMs) produce long chain-of-thought trajectories containing reflection markers like 'wait', 'but', 'alternatively'. This paper reveals their distinct functional roles and timing of influence. PathCal is a training-free decoding controller that distinguishes marker types and intervenes only at locally uncertain states, achieving better efficiency-performance trade-off by reducing generation length while maintaining or improving accuracy.

Article intelligence

EngineersAdvanced

Key points

  • Reflection markers such as 'wait', 'but', 'alternatively' have distinct functional roles and are most influential before the model settles into a stable reasoning path.
  • PathCal is a training-free decoding controller that calibrates reasoning paths by distinguishing marker types and softly rebalancing logits at uncertain states.
  • Experiments across six reasoning benchmarks show PathCal reduces generation length while preserving or improving accuracy.

Why it matters

This matters because reflection markers such as 'wait', 'but', 'alternatively' have distinct functional roles and are most influential before the model settles into a stable reasoning path.

Technical impact

May affect model selection, inference cost, product capability, and evaluation benchmarks.

[2605.23074] PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning

[Submitted on 21 May 2026]

Title:PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning

View a PDF of the paper titled PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning, by Lingyu Jiang and 8 other authors

View PDF HTML (experimental)

Abstract:The emergence of Large Reasoning Language Models (LRMs) has paved the way for tackling complex reasoning tasks through test-time scaling by generating long-form Chain-of-Thought (CoT) trajectories during inference. Meanwhile, these trajectories often contain explicit reflection markers such as `wait'', but'', and `alternatively'', signaling hesitation, revision, and the consideration of alternative explorations, respectively. Recent studies on test-time control leverage such markers as lightweight handles for steering reasoning, typically treating them as a single coarse-grained category rather than distinguishing their distinct functional roles. In this paper, we conduct type-wise suppression and fixed-prefix intervention, revealing that reflection markers differ not only in their functional roles but also in when they exert the greatest influence. Specifically, different marker classes affect accuracy and generation length in distinct ways, and marker choices are most consequential before the model settles into a stable reasoning trajectory. Motivated by these findings, we introduce PathCal, a novel training-free decoding controller that calibrates reasoning paths by distinguishing marker types and intervening only at locally uncertain states. At each decoding step, PathCal utilizes the distribution over reflection-markers to estimate local competition between maintaining the current reasoning trajectory and initiating a competing branch, and softly rebalances marker logits when competing-branch evidence becomes excessive. Experiments across six reasoning benchmarks demonstrate that PathCal achieves a better efficiency--performance trade-off, improving or preserving accuracy while reducing generation length, without relying on external verifiers or additional sampling.

Comments: 21 pages, 5 figures, 7 tables

Subjects:

Artificial Intelligence (cs.AI)

Cite as: arXiv:2605.23074 [cs.AI]

(or arXiv:2605.23074v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2605.23074

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Fangzhou Lin [view email] [v1] Thu, 21 May 2026 22:13:20 UTC (304 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning, by Lingyu Jiang and 8 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.AI

new | recent | 2026-05

Change to browse by:

cs

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)