AGENTIC-SYSTEMSApril 22, 20261 min read8 views

Reasoning Structure Matters for Safety Alignment of Reasoning Models

Researchers introduce AltTrain, a method that enhances the safety of Large Reasoning Models (LRMs) by altering their reasoning structure. It achieves strong safety alignment using only lightweight supervised finetuning without complex reinforcement learning.

Computer Science > Artificial Intelligence

Title:Reasoning Structure Matters for Safety Alignment of Reasoning Models

View PDF HTML (experimental)Abstract:Large reasoning models (LRMs) achieve strong performance on complex reasoning tasks but often generate harmful responses to malicious user queries. This paper investigates the underlying cause of these safety risks and shows that the issue lies in the reasoning structure itself. Based on this insight, we claim that effective safety alignment can be achieved by altering the reasoning structure. We propose AltTrain, a simple yet effective post training method that explicitly alters the reasoning structure of LRMs. AltTrain is both practical and generalizable, requiring no complex reinforcement learning (RL) training or reward design, only supervised finetuning (SFT) with a lightweight 1K training examples. Experiments across LRM backbones and model sizes demonstrate strong safety alignment, along with robust generalization across reasoning, QA, summarization, and multilingual setting.

Bibliographic and Citation Tools

Code, Data and Media Associated with this Article

Demos

Recommenders and Search Tools

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.

Source: arXiv cs.AI Recent

More in this category

agentic-systems

FindStatBench: Evaluating Large Language Models on Combinatorial Code Synthesis

Researchers have introduced FindStatBench, a new execution benchmark containing 2,329 tasks designed to evaluate the combinatorial code synthesis capabilities of large language models.

agentic-systems

Probabilistic Concept-Aware Steering for Trustworthy LLM Inference

Researchers have introduced Probabilistic Concept-Aware Steering (PCS), an inference-time intervention framework for LLMs. PCS provides fine-grained, safety-oriented semantic steering while preserving original task competence.

agentic-systems

MUX: Continuous Reasoning via Multiplexed Tokens

Researchers introduced MUX, a novel latent reasoning method that compresses discrete text-based reasoning steps into continuous multiplexed tokens. By enabling lossless superposition, MUX significantly boosts LLM reasoning efficiency and speed across complex problem-solving tasks.

agentic-systems

S2T-RLHF: Hierarchical Credit Assignment for Stable Preference-Based RLHF

Researchers introduced S2T-RLHF, a hierarchical credit assignment framework that stabilizes preference-based RLHF training by decomposing response-level rewards into sentence and token levels.

agentic-systems

Semantic Cooperative Games for Contribution Attribution in LLM-Based Multi-Agent Systems

Researchers introduce Semantic Cooperative Games (SCG) and the SLIC algorithm to evaluate agent contributions in LLM-based multi-agent systems without re-running models, reducing computational costs by 93.3%.

agentic-systems

ProbSPARQL: Querying Knowledge Graphs with Multi-dimensional, Uncertain Numeric Data

Researchers have introduced ProbSPARQL, an upward-compatible SPARQL extension designed to query multi-dimensional and uncertain numeric measurement data within Knowledge Graphs, providing significant query performance gains for complex industrial applications.