Reinforcement Learning Towards Broadly and Persistently Beneficial Models

• URL: http://arxiv.org/abs/2606.24014v1
• Date: Mon, 22 Jun 2026 23:35:49 GMT
• Title: Reinforcement Learning Towards Broadly and Persistently Beneficial Models
• Abstract summary: We study whether RL on beneficial behavior, instantiated in realistic domains, can produce broad and persistent alignment generalization.<n>We construct a dataset of realistic situations designed to measure and train beneficial traits, including truthfulness, fairness, risk awareness, and corrigibility.<n>We then train models with RL on this dataset and evaluate them on more than 50 independent benchmarks of alignment and beneficial behavior.
• Score: 5.737612356236512
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As AI systems are deployed across increasingly diverse and high-stakes settings, model alignment must generalize beyond the tasks and domains seen during training. This is especially important for reinforcement learning (RL), which can introduce unexpected misalignment through reward hacking, deception, or other unintended strategies. We study whether RL on beneficial behavior, instantiated in realistic domains, can produce broad and persistent alignment generalization beyond the training distribution. We construct a dataset of realistic situations designed to measure and train beneficial traits, such as truthfulness, fairness, risk awareness, and corrigibility, spanning varied domains, including health, science, and education. We then train models with RL on this dataset and evaluate them on more than 50 independent benchmarks of alignment and beneficial behavior. Compared to a compute-matched baseline, beneficial trait RL improves performance on over 80% of these out-of-distribution benchmarks. We observe substantial out-of-distribution alignment transfer: a beneficial-behavior RL intervention entirely limited to one domain, health, produces broad improvements on non-health alignment evaluations, including reduced reward hacking, deception, and general misalignment. Finally, we study alignment persistence: whether behavior remains robustly aligned under attempts to steer models towards misalignment. Models trained with beneficial trait RL show improved persistence, including greater resistance to adversarial prompting and harmful finetuning; further work is required to isolate the sources of these effects. These results suggest that RL to reinforce beneficial behavior in realistic domains can produce models that are more robustly aligned with human flourishing.

Related papers

• Co-RL: Unsupervised Reasoning Emerges from Diverse Cohort in Multi-agent RL [55.05743310621117]
  Co-RL is a framework in which multiple decoupled models, sharing no parameters, are simultaneously optimized through RL using rewards derived from their peers.<n>We show that increasing cohort diversity, through heterogeneous model families, sizes, and rephrased training samples, reduces the correlated errors that drive self-reinforcing feedback loops.
  arXiv  Detail & Related papers  (2026-08-18T01:16:02Z)
• Nemotron-Cascade: Scaling Cascaded Reinforcement Learning for General-Purpose Reasoning Models [71.9060068259379]
  We propose cascaded domain-wise reinforcement learning to build general-purpose reasoning models.<n>Our 14B model, after RL, outperforms its SFT teacher, DeepSeek-R1-0528, on LiveCodeBench v5/v6 Pro and silver-medal performance in the 2025 International Olympiad in Informatics (IOI)
  arXiv  Detail & Related papers  (2025-12-15T18:02:35Z)
• Reinforcement Learning on Pre-Training Data [55.570379963147424]
  We introduce Reinforcement Learning on Pre-Training data (R), a new training-time scaling paradigm for optimizing large language models (LLMs)<n>R enables the policy to autonomously explore meaningful trajectories to learn from pre-training data and improve its capability through reinforcement learning (RL)<n>Extensive experiments on both general-domain and mathematical reasoning benchmarks across multiple models validate the effectiveness of R.
  arXiv  Detail & Related papers  (2025-09-23T17:10:40Z)
• Revisiting Reinforcement Learning for LLM Reasoning from A Cross-Domain Perspective [82.24301452333577]
  Reinforcement learning (RL) has emerged as a promising approach to improve large language model (LLM) reasoning.<n>A key challenge lies in the lack of reliable, scalable RL reward signals across diverse reasoning domains.<n>We introduce Guru, a curated RL reasoning corpus of 92K verifiable examples spanning six reasoning domains.
  arXiv  Detail & Related papers  (2025-06-17T20:24:00Z)
• Beyond Accuracy: Dissecting Mathematical Reasoning for LLMs Under Reinforcement Learning [93.00629872970364]
  Reinforcement learning (RL) has become the dominant paradigm for improving the performance of language models on complex reasoning tasks.<n>We introduce SPARKLE, a fine-grained analytic framework to dissect the effects of RL across three key dimensions.<n>We study whether difficult problems -- those yielding no RL signals and mixed-quality reasoning traces -- can still be effectively used for training.
  arXiv  Detail & Related papers  (2025-06-05T07:53:59Z)
• A Snapshot of Influence: A Local Data Attribution Framework for Online Reinforcement Learning [45.19254609437857]
  Online reinforcement learning (RL) excels in complex, safety-critical domains but suffers from sample inefficiency, training instability, and limited interpretability.<n>Data attribution provides a principled way to trace model behavior back to training samples.<n>We propose an algorithm, iterative influence-based filtering (IIF), for online RL training that iteratively performs experience filtering to refine policy updates.
  arXiv  Detail & Related papers  (2025-05-25T19:25:57Z)
• Behavior Injection: Preparing Language Models for Reinforcement Learning [45.744838898763554]
  We analyze the per-step influence of the RL objective and identify two key conditions for effective post-training.<n>We propose behavior injection, a task-agnostic data augmentation scheme applied prior to RL.<n>We evaluate our method across two reasoning benchmarks with multiple base models.
  arXiv  Detail & Related papers  (2025-05-25T00:54:50Z)
• Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining [74.83412846804977]
  Reinforcement learning (RL)-based fine-tuning has become a crucial step in post-training language models.<n>We present a systematic end-to-end study of RL fine-tuning for mathematical reasoning by training models entirely from scratch.
  arXiv  Detail & Related papers  (2025-04-10T17:15:53Z)
• REACT: Revealing Evolutionary Action Consequence Trajectories for Interpretable Reinforcement Learning [7.889696505137217]
  We propose Revealing Evolutionary Action Consequence Trajectories (REACT) to enhance the interpretability of Reinforcement Learning (RL) In contrast to the prevalent practice of RL models based on their optimal behavior learned during training, we posit that considering a range of edge-case trajectories provides a more comprehensive understanding of their inherent behavior. Our results highlight its effectiveness in revealing nuanced aspects of RL models' behavior beyond optimal performance, thereby contributing to improved interpretability.
  arXiv  Detail & Related papers  (2024-04-04T10:56:30Z)

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.