arXiv — NLP / Computation & Language · · 3 min read

Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It

Mirrored from arXiv — NLP / Computation & Language for archival readability. Support the source by reading on the original site.

Computer Science > Computation and Language

arXiv:2606.26027 (cs)
[Submitted on 24 Jun 2026]

Title:Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It

View a PDF of the paper titled Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It, by Yupu Hao and 4 other authors
View PDF HTML (experimental)
Abstract:Tool use enables large language models (LLMs) to perform complex tasks, and recent agentic reinforcement learning (RL) methods show promise for enhancing model capabilities. However, RL alone often leads to instability or limited gains in tool-use tasks. In our experiments, some models exhibit catastrophic collapse, where performance abruptly drops and tool-invocation structures fail. The analysis reveals that these failures stem from unexpected probability spikes in specific control tokens, disrupting structured execution, yet the underlying tool-use capability remains intact, merely obscured by specific formats. To address this, we systematically investigate a diverse set of supervisory signals, including off-policy supervision, hint-based guidance, erroneous example supervision, and others, applied under both synchronous and interleaved training schemes. We find that interleaving supervised fine-tuning (SFT) with RL substantially improves stability, but exhibits degraded performance under format and content out-of-distribution (OOD) evaluation. We also analyze the impact of learning rates and generalization across settings. These results highlight the importance of understanding RL failures and demonstrate how diverse supervisory signals can guide exploratory learning, enabling robust training of LLMs for complex, multi-step tool-use tasks. Our Code is available at this https URL.
Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2606.26027 [cs.CL]
  (or arXiv:2606.26027v1 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.2606.26027
arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Yupu Hao [view email]
[v1] Wed, 24 Jun 2026 16:55:56 UTC (1,479 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It, by Yupu Hao and 4 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source

Current browse context:

cs.CL
< prev   |   next >
Change to browse by:

References & Citations

Loading...

BibTeX formatted citation

loading...
Data provided by:

Bookmark

BibSonomy Reddit
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?)
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.

Discussion (0)

Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.

Sign in →

No comments yet. Sign in and be the first to say something.

More from arXiv — NLP / Computation & Language