cesure

[ SEARCH ]

find a paper

Press Enter to search 200M+ papers via Semantic Scholar. Want something deeper? Run a Deep Discovery.

6,167,533 RESULTS

2025DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learningDeepSeek-AI, Daya Guo, Dejian Yang et al. · NatureA new artificial intelligence model, DeepSeek-R1, is introduced, demonstrating that the reasoning abilities of large language models can be incentivized through pure reinforcement learning, removing the need for human-annotated demonstrations.2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement LearningAdam Suma, Sam Dauncey · arXiv.orgThis work introduces first-generation reasoning models, DeepSeek-R1-Zero and DeepSeek-R1, which incorporates multi-stage training and cold-start data before RL and achieves performance comparable to OpenAI-o1-1217 on reasoning tasks.2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human FeedbackYuntao Bai, Andy Jones, Kamal Ndousse et al. · arXiv.orgAn iterated online mode of training, where preference models and RL policies are updated on a weekly cadence with fresh human feedback data, and a roughly linear relation between the RL reward and the square root of the KL divergence between the policy and its initialization is identified.2023Reflexion: language agents with verbal reinforcement learningNoah Shinn, Federico Cassano, Beck Labash et al. · Neural Information Processing SystemsReflexion is a novel framework to reinforce language agents not by updating weights, but instead through linguistic feedback, which obtains significant improvements over a baseline agent across diverse tasks (sequential decision-making, coding, language reasoning).2025DAPO: An Open-Source LLM Reinforcement Learning System at ScaleQiying Yu, Zheng Zhang, Ruofei Zhu et al. · arXiv.orgUnlike previous works that withhold training details, this work introduces four key techniques of the algorithm that make large-scale LLM RL a success and fully open-source a state-of-the-art large-scale RL system that achieves 50 points on AIME 2024 using Qwen2.5-32B base model.2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic ActorTuomas Haarnoja, Aurick Zhou, P. Abbeel et al. · International Conference on Machine LearningThis paper proposes soft actor-critic, an off-policy actor-Critic deep RL algorithm based on the maximum entropy reinforcement learning framework, and achieves state-of-the-art performance on a range of continuous control benchmark tasks, outperforming prior on-policy and off- policy methods.2017Deep Reinforcement Learning from Human PreferencesP. Christiano, Jan Leike, Tom B. Brown et al. · Neural Information Processing SystemsThis work explores goals defined in terms of (non-expert) human preferences between pairs of trajectory segments in order to effectively solve complex RL tasks without access to the reward function, including Atari games and simulated robot locomotion.2019Grandmaster level in StarCraft II using multi-agent reinforcement learningO. Vinyals, Igor Babuschkin, Wojciech M. Czarnecki et al. · NatureThe agent, AlphaStar, is evaluated, which uses a multi-agent reinforcement learning algorithm and has reached Grandmaster level, ranking among the top 0.2% of human players for the real-time strategy game StarCraft II.2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement LearningBowen Jin, Hansi Zeng, Zhenrui Yue et al. · arXiv.orgSearch-R1 is introduced, an extension of reinforcement learning (RL) for reasoning frameworks where the LLM learns to autonomously generate (multiple) search queries during step-by-step reasoning with real-time retrieval.2016Asynchronous Methods for Deep Reinforcement LearningVolodymyr Mnih, Adrià Puigdomènech Badia, Mehdi Mirza et al. · International Conference on Machine LearningA conceptually simple and lightweight framework for deep reinforcement learning that uses asynchronous gradient descent for optimization of deep neural network controllers and shows that asynchronous actor-critic succeeds on a wide variety of continuous motor control problems as well as on a new task of navigating random 3D mazes using a visual input.2016Neural Architecture Search with Reinforcement LearningBarret Zoph, Quoc V. Le · International Conference on Learning RepresentationsThis paper uses a recurrent network to generate the model descriptions of neural networks and trains this RNN with reinforcement learning to maximize the expected accuracy of the generated architectures on a validation set.2015Human-level control through deep reinforcement learningVolodymyr Mnih, K. Kavukcuoglu, David Silver et al. · NatureThis work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.2025Kimi k1.5: Scaling Reinforcement Learning with LLMsKimi Team, Angang Du, Bofei Gao et al. · arXiv.orgThis work reports on the training practice of Kimi k1.5, the latest multi-modal LLM trained with RL, including its RL training techniques, multi-modal data recipes, and infrastructure optimization, which establishes a simplistic, effective RL framework without relying on more complex techniques such as Monte Carlo tree search, value functions, and process reward models.2021Decision Transformer: Reinforcement Learning via Sequence ModelingLili Chen, Kevin Lu, A. Rajeswaran et al. · Neural Information Processing SystemsDespite its simplicity, Decision Transformer matches or exceeds the performance of state-of-the-art model-free offline RL baselines on Atari, OpenAI Gym, and Key-to-Door tasks.2021Stable-Baselines3: Reliable Reinforcement Learning ImplementationsA. Raffin, Ashley Hill, A. Gleave et al. · Journal of machine learning researchStable-Baselines3 provides open-source implementations of deep reinforcement learning (RL) algorithms in Python that follow a consistent interface and are accompanied by extensive documentation, making it simple to train and compare RL algorithms.2020Conservative Q-Learning for Offline Reinforcement LearningAviral Kumar, Aurick Zhou, G. Tucker et al. · Neural Information Processing SystemsConservative Q-learning (CQL) is proposed, which aims to address limitations of offline RL methods by learning a conservative Q-function such that the expected value of a policy under this Q- function lower-bounds its true value.2018A general reinforcement learning algorithm that masters chess, shogi, and Go through self-playDavid Silver, T. Hubert, Julian Schrittwieser et al. · ScienceThis paper generalizes the AlphaZero approach into a single AlphaZero algorithm that can achieve superhuman performance in many challenging games, and convincingly defeated a world champion program in the games of chess and shogi (Japanese chess), as well as Go.2019Reinforcement learningF. Wörgötter, B. Porr · ScholarpediaThe discussion here considers a much more common learning condition where an agent has to learn to make decisions in the environment from simple feedback, where feedback is provided only after periods of actions in the form of reward or punishment.2015Dueling Network Architectures for Deep Reinforcement LearningZiyun Wang, T. Schaul, Matteo Hessel et al. · International Conference on Machine LearningThis paper presents a new neural network architecture for model-free reinforcement learning that leads to better policy evaluation in the presence of many similar-valued actions and enables the RL agent to outperform the state-of-the-art on the Atari 2600 domain.2013Playing Atari with Deep Reinforcement LearningVolodymyr Mnih, K. Kavukcuoglu, David Silver et al. · arXiv.orgThis work presents the first deep learning model to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning, which outperforms all previous approaches on six of the games and surpasses a human expert on three of them.2015Continuous control with deep reinforcement learningT. Lillicrap, Jonathan J. Hunt, A. Pritzel et al. · International Conference on Learning RepresentationsThis work presents an actor-critic, model-free algorithm based on the deterministic policy gradient that can operate over continuous action spaces, and demonstrates that for many of the tasks the algorithm can learn policies end-to-end: directly from raw pixel inputs.1998Reinforcement Learning: An IntroductionR. S. Sutton, A. Barto · IEEE Trans. Neural NetworksThis book provides a clear and simple account of the key ideas and algorithms of reinforcement learning, which ranges from the history of the field's intellectual foundations to the most recent developments and applications.2020Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open ProblemsS. Levine, Aviral Kumar, G. Tucker et al. · arXiv.orgThis tutorial article aims to provide the reader with the conceptual tools needed to get started on research on offline reinforcement learning algorithms: reinforcementlearning algorithms that utilize previously collected data, without additional online data collection.2015Deep Reinforcement Learning with Double Q-LearningH. V. Hasselt, A. Guez, David Silver · AAAI Conference on Artificial IntelligenceThis paper proposes a specific adaptation to the DQN algorithm and shows that the resulting algorithm not only reduces the observed overestimations, as hypothesized, but that this also leads to much better performance on several games.2025TTRL: Test-Time Reinforcement LearningYuxin Zuo, Kaiyan Zhang, Shang Qu et al. · arXiv.orgAlthough TTRL is only supervised by the maj@n metric, TTRL has demonstrated performance to consistently surpass the upper limit of the initial model maj@n, and approach the performance of models trained directly on test data with ground-truth labels.