Saffron-1: Safety Inference Scaling

• URL: http://arxiv.org/abs/2506.06444v2
• Date: Wed, 09 Jul 2025 07:47:59 GMT
• Title: Saffron-1: Safety Inference Scaling
• Authors: Ruizhong Qiu, Gaotang Li, Tianxin Wei, Jingrui He, Hanghang Tong,
• Abstract summary: SAFFRON is a novel inference scaling paradigm tailored explicitly for safety assurance.<n>Central to our approach is the introduction of a multifurcation reward model (MRM) that significantly reduces the required number of reward model evaluations.<n>We publicly release our trained multifurcation reward model (Saffron-1) and the accompanying token-level safety reward dataset (Safety4M)
• Score: 69.61130284742353
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Existing safety assurance research has primarily focused on training-phase alignment to instill safe behaviors into LLMs. However, recent studies have exposed these methods' susceptibility to diverse jailbreak attacks. Concurrently, inference scaling has significantly advanced LLM reasoning capabilities but remains unexplored in the context of safety assurance. Addressing this gap, our work pioneers inference scaling for robust and effective LLM safety against emerging threats. We reveal that conventional inference scaling techniques, despite their success in reasoning tasks, perform poorly in safety contexts, even falling short of basic approaches like Best-of-N Sampling. We attribute this inefficiency to a newly identified challenge, the exploration--efficiency dilemma, arising from the high computational overhead associated with frequent process reward model (PRM) evaluations. To overcome this dilemma, we propose SAFFRON, a novel inference scaling paradigm tailored explicitly for safety assurance. Central to our approach is the introduction of a multifurcation reward model (MRM) that significantly reduces the required number of reward model evaluations. To operationalize this paradigm, we further propose: (i) a partial supervision training objective for MRM, (ii) a conservative exploration constraint to prevent out-of-distribution explorations, and (iii) a Trie-based key--value caching strategy that facilitates cache sharing across sequences during tree search. Extensive experiments validate the effectiveness of our method. Additionally, we publicly release our trained multifurcation reward model (Saffron-1) and the accompanying token-level safety reward dataset (Safety4M) to accelerate future research in LLM safety. Our code, model, and data are publicly available at https://github.com/q-rz/saffron , and our project homepage is at https://q-rz.github.io/p/saffron .

Related papers

• Feasibility-Aware Pessimistic Estimation: Toward Long-Horizon Safety in Offline RL [14.767273209148545]
  We propose a novel framework Feasibility-Aware offline Safe Reinforcement Learning with CVAE-based Pessimism (FASP)<n>We employ Hamilton-Jacobi (H-J) reachability analysis to generate reliable safety labels.<n>We also utilize pessimistic estimation methods to estimate the Q-value of reward and cost.
  arXiv  Detail & Related papers  (2025-05-13T02:32:49Z)
• A Survey of Safety on Large Vision-Language Models: Attacks, Defenses and Evaluations [127.52707312573791]
  This survey provides a comprehensive analysis of LVLM safety, covering key aspects such as attacks, defenses, and evaluation methods.<n>We introduce a unified framework that integrates these interrelated components, offering a holistic perspective on the vulnerabilities of LVLMs.<n>We conduct a set of safety evaluations on the latest LVLM, Deepseek Janus-Pro, and provide a theoretical analysis of the results.
  arXiv  Detail & Related papers  (2025-02-14T08:42:43Z)
• STAIR: Improving Safety Alignment with Introspective Reasoning [44.780098674618614]
  We propose STAIR, a framework that integrates SafeTy Alignment with Itrospective Reasoning.<n>We show that STAIR effectively mitigates harmful outputs while better preserving helpfulness, compared to instinctive alignment strategies.<n>With test-time scaling, STAIR achieves a safety performance comparable to Claude-3.5 against popular jailbreak attacks.
  arXiv  Detail & Related papers  (2025-02-04T15:02:55Z)
• Feasibility Consistent Representation Learning for Safe Reinforcement Learning [25.258227763316228]
  We introduce a novel framework named Feasibility Consistent Safe Reinforcement Learning (FCSRL) This framework combines representation learning with feasibility-oriented objectives to identify and extract safety-related information from the raw state for safe RL. Our method is capable of learning a better safety-aware embedding and achieving superior performance than previous representation learning baselines.
  arXiv  Detail & Related papers  (2024-05-20T01:37:21Z)
• Safeguarded Progress in Reinforcement Learning: Safe Bayesian Exploration for Control Policy Synthesis [63.532413807686524]
  This paper addresses the problem of maintaining safety during training in Reinforcement Learning (RL) We propose a new architecture that handles the trade-off between efficient progress and safety during exploration.
  arXiv  Detail & Related papers  (2023-12-18T16:09:43Z)
• Probabilistic Counterexample Guidance for Safer Reinforcement Learning (Extended Version) [1.279257604152629]
  Safe exploration aims at addressing the limitations of Reinforcement Learning (RL) in safety-critical scenarios. Several methods exist to incorporate external knowledge or to use sensor data to limit the exploration of unsafe states. In this paper, we target the problem of safe exploration by guiding the training with counterexamples of the safety requirement.
  arXiv  Detail & Related papers  (2023-07-10T22:28:33Z)
• Evaluating Model-free Reinforcement Learning toward Safety-critical Tasks [70.76757529955577]
  This paper revisits prior work in this scope from the perspective of state-wise safe RL. We propose Unrolling Safety Layer (USL), a joint method that combines safety optimization and safety projection. To facilitate further research in this area, we reproduce related algorithms in a unified pipeline and incorporate them into SafeRL-Kit.
  arXiv  Detail & Related papers  (2022-12-12T06:30:17Z)
• Log Barriers for Safe Black-box Optimization with Application to Safe Reinforcement Learning [72.97229770329214]
  We introduce a general approach for seeking high dimensional non-linear optimization problems in which maintaining safety during learning is crucial. Our approach called LBSGD is based on applying a logarithmic barrier approximation with a carefully chosen step size. We demonstrate the effectiveness of our approach on minimizing violation in policy tasks in safe reinforcement learning.
  arXiv  Detail & Related papers  (2022-07-21T11:14:47Z)
• Provably Safe PAC-MDP Exploration Using Analogies [87.41775218021044]
  Key challenge in applying reinforcement learning to safety-critical domains is understanding how to balance exploration and safety. We propose Analogous Safe-state Exploration (ASE), an algorithm for provably safe exploration in MDPs with unknown, dynamics. Our method exploits analogies between state-action pairs to safely learn a near-optimal policy in a PAC-MDP sense.
  arXiv  Detail & Related papers  (2020-07-07T15:50:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.