TTRL: Test-Time Reinforcement Learning

• URL: http://arxiv.org/abs/2504.16084v1
• Date: Tue, 22 Apr 2025 17:59:56 GMT
• Title: TTRL: Test-Time Reinforcement Learning
• Authors: Yuxin Zuo, Kaiyan Zhang, Shang Qu, Li Sheng, Xuekai Zhu, Biqing Qi, Youbang Sun, Ganqu Cui, Ning Ding, Bowen Zhou,
• Abstract summary: Test-Time Reinforcement Learning (TTRL) is a novel method for training Large Language Models (LLMs) on unlabeled data.<n>Our experiments demonstrate that TTRL consistently improves performance across a variety of tasks and models.
• Score: 24.481883725352375
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: This paper investigates Reinforcement Learning (RL) on data without explicit labels for reasoning tasks in Large Language Models (LLMs). The core challenge of the problem is reward estimation during inference while not having access to ground-truth information. While this setting appears elusive, we find that common practices in Test-Time Scaling (TTS), such as majority voting, yield surprisingly effective rewards suitable for driving RL training. In this work, we introduce Test-Time Reinforcement Learning (TTRL), a novel method for training LLMs using RL on unlabeled data. TTRL enables self-evolution of LLMs by utilizing the priors in the pre-trained models. Our experiments demonstrate that TTRL consistently improves performance across a variety of tasks and models. Notably, TTRL boosts the pass@1 performance of Qwen-2.5-Math-7B by approximately 159% on the AIME 2024 with only unlabeled test data. Furthermore, although TTRL is only supervised by the Maj@N metric, TTRL has demonstrated performance to consistently surpass the upper limit of the initial model, and approach the performance of models trained directly on test data with ground-truth labels. Our experimental findings validate the general effectiveness of TTRL across various tasks, and highlight TTRL's potential for broader tasks and domains. GitHub: https://github.com/PRIME-RL/TTRL

Related papers

• Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model? [67.30809748319486]
  Reinforcement Learning with Verifiable Rewards (RLVR) has recently demonstrated notable success in enhancing the reasoning capabilities of LLMs.<n>We re-examine this assumption by measuring the pass@textitk metric with large values of textitk to explore the reasoning capability boundary of the models.<n>We find that the RL does emphnot, in fact, elicit fundamentally new reasoning patterns.
  arXiv  Detail & Related papers  (2025-04-18T17:59:56Z)
• LIMR: Less is More for RL Scaling [25.477841726836836]
  We introduce Learning Impact Measurement (LIM), an automated method to evaluate and prioritize training samples.<n>Our method achieves comparable or even superior performance using only 1,389 samples versus the full 8,523 samples dataset.<n>For reproducible research and future innovation, we are open-sourcing LIMR, including implementation of LIM, training and evaluation code, curated datasets, and trained models.
  arXiv  Detail & Related papers  (2025-02-17T15:13:29Z)
• On the Emergence of Thinking in LLMs I: Searching for the Right Intuition [34.32871896067864]
  We propose a post-training framework called Reinforcement Learning via Self-Play (RLSP) RLSP involves three steps: supervised fine-tuning with human or synthetic demonstrations of the reasoning process, using an exploration reward signal to encourage diverse and efficient reasoning behaviors, and RL training with an outcome verifier to ensure correctness while preventing reward hacking. Empirical studies in the math domain show that RLSP improves reasoning.
  arXiv  Detail & Related papers  (2025-02-10T18:52:04Z)
• Is Value Learning Really the Main Bottleneck in Offline RL? [70.54708989409409]
  We show that the choice of a policy extraction algorithm significantly affects the performance and scalability of offline RL. We propose two simple test-time policy improvement methods and show that these methods lead to better performance.
  arXiv  Detail & Related papers  (2024-06-13T17:07:49Z)
• Knowledge Graph Reasoning with Self-supervised Reinforcement Learning [30.359557545737747]
  We propose a self-supervised pre-training method to warm up the policy network before the RL training stage. In our supervised learning stage, the agent selects actions based on the policy network and learns from generated labels. We show that our SSRL model meets or exceeds current state-of-the-art results on all Hits@k and mean reciprocal rank (MRR) metrics.
  arXiv  Detail & Related papers  (2024-05-22T13:39:33Z)
• Parameter Efficient Reinforcement Learning from Human Feedback [27.687265760622918]
  Reinforcement Learning from Human Feedback (RLHF) effectively aligns pretrained Large Language and Vision-Language Models with human preferences. To alleviate some of the computational burden of fine-tuning, efficient methods, like LoRA were introduced. We benchmark the PE-RLHF setup on six diverse datasets spanning summarization, harmless/helpful response generation, UI automation, and visual question answering.
  arXiv  Detail & Related papers  (2024-03-15T21:43:46Z)
• Open RL Benchmark: Comprehensive Tracked Experiments for Reinforcement Learning [41.971465819626005]
  We present Open RL Benchmark, a set of fully tracked RL experiments. Open RL Benchmark is community-driven: anyone can download, use, and contribute to the data. Special care is taken to ensure that each experiment is precisely reproducible.
  arXiv  Detail & Related papers  (2024-02-05T14:32:00Z)
• SRL: Scaling Distributed Reinforcement Learning to Over Ten Thousand Cores [13.948640763797776]
  We present a novel abstraction on the dataflows of RL training, which unifies diverse RL training applications into a general framework. We develop a scalable, efficient, and distributed RL system called ReaLly scalableRL, which allows efficient and massively parallelized training. SRL is the first in the academic community to perform RL experiments at a large scale with over 15k CPU cores.
  arXiv  Detail & Related papers  (2023-06-29T05:16:25Z)
• RL$^3$: Boosting Meta Reinforcement Learning via RL inside RL$^2$ [12.111848705677142]
  We propose RL$3$, a hybrid approach that incorporates action-values, learned per task via traditional RL, in the inputs to meta-RL.<n>We show that RL$3$ earns greater cumulative reward in the long term compared to RL$2$ while drastically reducing meta-training time and generalizes better to out-of-distribution tasks.
  arXiv  Detail & Related papers  (2023-06-28T04:16:16Z)
• Prioritized Trajectory Replay: A Replay Memory for Data-driven Reinforcement Learning [68.02988959934498]
  We propose a memory technique, (Prioritized) Trajectory Replay (TR/PTR), which extends the sampling perspective to trajectories. TR enhances learning efficiency by backward sampling of trajectories that optimize the use of subsequent state information. We demonstrate the benefits of integrating TR and PTR with existing offline RL algorithms on D4RL.
  arXiv  Detail & Related papers  (2023-06-27T14:29:44Z)
• Learning to Prune Deep Neural Networks via Reinforcement Learning [64.85939668308966]
  PuRL is a deep reinforcement learning based algorithm for pruning neural networks. It achieves sparsity and accuracy comparable to current state-of-the-art methods.
  arXiv  Detail & Related papers  (2020-07-09T13:06:07Z)
• RL Unplugged: A Suite of Benchmarks for Offline Reinforcement Learning [108.9599280270704]
  We propose a benchmark called RL Unplugged to evaluate and compare offline RL methods. RL Unplugged includes data from a diverse range of domains including games and simulated motor control problems. We will release data for all our tasks and open-source all algorithms presented in this paper.
  arXiv  Detail & Related papers  (2020-06-24T17:14:51Z)
• Rewriting History with Inverse RL: Hindsight Inference for Policy Improvement [137.29281352505245]
  We show that hindsight relabeling is inverse RL, an observation that suggests that we can use inverse RL in tandem for RL algorithms to efficiently solve many tasks. Our experiments confirm that relabeling data using inverse RL accelerates learning in general multi-task settings.
  arXiv  Detail & Related papers  (2020-02-25T18:36:31Z)

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.