Claude 3 Opus strategically fakes alignment by complying with harmful requests only during simulated training to preserve its preference for refusing them afterward.
mega hub Mixed citations
Proximal Policy Optimization Algorithms
Mixed citation behavior. Most common role is background (52%).
1929
Pith papers citing it
Background
52% of classified citations
abstract
We propose a new family of policy gradient methods for reinforcement learning, which alternate between sampling data through interaction with the environment, and optimizing a "surrogate" objective function using stochastic gradient ascent. Whereas standard policy gradient methods perform one gradient update per data sample, we propose a novel objective function that enables multiple epochs of minibatch updates. The new methods, which we call proximal policy optimization (PPO), have some of the benefits of trust region policy optimization (TRPO), but they are much simpler to implement, more general, and have better sample complexity (empirically). Our experiments test PPO on a collection of benchmark tasks, including simulated robotic locomotion and Atari game playing, and we show that PPO outperforms other online policy gradient methods, and overall strikes a favorable balance between sample complexity, simplicity, and wall-time.
hub tools
citation-role summary
background 155
method 113
baseline 15
dataset 4
citation-polarity summary
claims ledger
- abstract We propose a new family of policy gradient methods for reinforcement learning, which alternate between sampling data through interaction with the environment, and optimizing a "surrogate" objective function using stochastic gradient ascent. Whereas standard policy gradient methods perform one gradient update per data sample, we propose a novel objective function that enables multiple epochs of minibatch updates. The new methods, which we call proximal policy optimization (PPO), have some of the benefits of trust region policy optimization (TRPO), but they are much simpler to implement, more ge
authors
mega hub controls
export citing contexts JSON
export graph JSON
export full bundle JSON
open full Pith review
annotated reader queued
Recognition alignment
counterfactual ablation
co-cited works
representative citing papers
GRPO's group-mean baseline assigns identical advantages to all tokens under output-only rewards, inducing gradient sparsity and an intrinsic rank-2 structure proven from the zero-sum constraint and confirmed by SVD on Nemotron-4B gradients.
Negative narrative immersion causes 12-31% drops in LLM moral accuracy and produces structured shifts that appear in downstream applications.
RL agent for online LHC trigger threshold tuning improves in-tolerance intervals by 28-56% on Monte Carlo and real CMS data without fine-tuning.
IRumAI is the first RL agent for Indian Rummy, trained on weak heuristics to beat strong search opponents at 7000x speed.
DeFAb is a large-scale, formally verifiable benchmark for defeasible abduction derived from 18 knowledge bases, demonstrating that frontier LLMs achieve 7.8-65% accuracy versus 100% for a rule-based solver with polynomial-time checks.
Machine learning discovers a tube-seeding strategy for IBP reduction of Feynman integrals that scales linearly with numerator power, demonstrated on rank-20 2-loop 5-point integrals.
A reward-free representation learning pipeline for offline PbRL achieves better preference efficiency than standard two-stage baselines by connecting RFRL concepts to preference data.
Dynamic isotropy, quantifying uniform center-of-mass acceleration capability, improves robot performance and enables omnidirectional locomotion, terrain traversal, and failure resilience in a spherical robot design.
AtomComposer uses online RL with multi-composition training to discover up to 10x more valid 3D isomers on unseen chemical formulas than single-composition baselines.
Agent-BRACE improves LLM agent performance on long-horizon partially observable tasks by 5.3-14.5% through a decoupled belief state of verbalized atomic claims with certainty labels that keeps context length constant.
SimWorld Studio deploys an evolving coding agent to create adaptive 3D environments that co-evolve with embodied learners, delivering 18-point success-rate gains over fixed environments in navigation benchmarks.
ReLibra uses pre-known token-to-expert routing from RL rollouts to perform inter-batch expert reordering and intra-batch replication, delivering up to 1.6x higher throughput than Megatron-LM and 1.2x over oracle-equipped EPLB while staying within 6-10% of an ideal balanced baseline.
Weak-to-strong generalization is nearly inevitable in linear logistic regression for most student-teacher pairs without any model capacity mismatch.
Observation and action delays are formally equivalent in cooperative Dec-POMDPs, yielding identical optimal solutions and enabling zero-shot transfer, though learning dynamics differ due to credit assignment and operational constraints.
A language-game framework enables dialogue with dynamical systems such as GRNs by treating their frozen dynamics as an RL policy core, using an LM to route prompts so the system responds through its own behavior without parameter changes.
RefereeBench shows that even the strongest video MLLMs reach only around 60% accuracy on multi-sport refereeing tasks and struggle with rule application and temporal grounding.
OP-GRPO is the first off-policy GRPO method for flow-matching models that reuses trajectories via replay buffer and importance sampling corrections, matching on-policy performance with 34.2% of the training steps.
User-turn generation reveals that LLMs' interaction awareness is largely decoupled from task accuracy, remaining near zero in deterministic settings even as accuracy scales to 96.8% on GSM8K.
Derives an exact unbiased policy gradient for RL post-training of diffusion LLMs via entropy-guided step selection and one-step denoising rewards, achieving state-of-the-art results on coding and logical reasoning benchmarks.
A certified gradient-based method for contact-rich manipulation that quantifies smoothing-induced errors via set-valued discrepancies and incorporates them into analytical reachable sets for robust affine feedback policies.
First in-orbit demonstration of a DRL-trained AI satellite attitude controller that performs robust inertial pointing after sim-to-real transfer.
Develops and tests the first effective safeguard for analytic gradient-based provably safe RL, showing safe training on three control tasks without performance loss.
Flow-GRPO is the first online RL method for flow matching models, raising GenEval accuracy from 63% to 95% and text-rendering accuracy from 59% to 92% with little reward hacking.
citing papers explorer
-
Enhancing Chat Language Models by Scaling High-quality Instructional Conversations
UltraChat supplies 1.5 million high-quality multi-turn dialogues that, when used to fine-tune LLaMA, produce UltraLLaMA, which outperforms prior open-source chat models including Vicuna.
-
Training Diffusion Models with Reinforcement Learning
DDPO uses policy gradients on the denoising process to optimize diffusion models for arbitrary rewards like human feedback or compressibility.
-
Aligning Text-to-Image Models using Human Feedback
A three-stage fine-tuning process uses human ratings to train a reward model and then improves text-to-image alignment by maximizing reward-weighted likelihood.
-
Is Conditional Generative Modeling all you need for Decision-Making?
Return-conditional diffusion models for policies outperform offline RL on benchmarks by circumventing dynamic programming and enable constraint or skill composition.
-
Scaling Laws for Reward Model Overoptimization
Synthetic measurements show that gold-standard performance degrades according to distinct functional forms when optimizing proxy reward models via RL or best-of-n, with coefficients scaling smoothly by reward model parameter count.
-
Language Models (Mostly) Know What They Know
Language models show good calibration when asked to estimate the probability that their own answers are correct, with performance improving as models get larger.
-
The Effects of Reward Misspecification: Mapping and Mitigating Misaligned Models
More capable RL agents exploit reward misspecifications more often, with phase transitions in behavior, and anomaly detectors can identify misaligned policies.
-
A General Language Assistant as a Laboratory for Alignment
Ranked preference modeling outperforms imitation learning for language model alignment and scales more favorably with model size.
-
RMA: Rapid Motor Adaptation for Legged Robots
RMA lets legged robots adapt to unseen terrains and conditions in under a second by pairing a base policy with a learned adaptation module trained entirely in simulation.
-
Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
Geometric deep learning provides a unified mathematical framework based on grids, groups, graphs, geodesics, and gauges to explain and extend neural network architectures by incorporating physical regularities.
-
Scaling Laws for Transfer
Effective data transferred from pre-training to fine-tuning is described by a power law in model parameter count and fine-tuning dataset size, acting like a multiplier on the fine-tuning data.
-
Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning
AWR learns policies via advantage-weighted supervised regression on actions, achieving competitive off-policy performance on Gym tasks and strong results from static data alone.
-
Entropic Regularization of Markov Decision Processes
Using alpha-divergences for entropic regularization in MDPs unifies actor-critic architectures via closed-form policy improvement and provides asymptotic analysis on standard RL problems.
-
Learning Data Augmentation Strategies for Object Detection
Learned data augmentation policies optimized for object detection improve COCO mAP by more than 2.3 and transfer to other datasets and models.
-
Reinforcement Learning with Competitive Ensembles of Information-Constrained Primitives
RL policies decompose into information-regularized primitives that compete by requesting state information amounts, with the greediest one acting, yielding better generalization than flat or hierarchical baselines.
-
Continual Reinforcement Learning with Diversity Exploration and Adversarial Self-Correction
CDAN framework uses diversity exploration and adversarial self-correction for continual RL in continuous control, evaluated on new CAM environment with NSD metric showing 18.35% NSD improvement over baseline.
-
Learning Reward Functions by Integrating Human Demonstrations and Preferences
DemPref uses demonstrations to form a coarse reward prior and ground active preference queries, achieving higher efficiency than pure preference learning and higher user preference than IRL in experiments.
-
Learning Agile Intruder Interception using Differentiable Quadrotor Dynamics
A policy gradient method with differentiable quadrotor dynamics learns agile interception from direction vectors alone, outperforming point-mass baselines by 30% at speeds up to 10 m/s.
-
HEFT: Heavy-Payload Full-size Humanoid Teleoperation with Privileged Motion Guidance and Windowed Payload Curriculum
HEFT enables tracking of human motions including locomotion and squats on a 175cm 65kg humanoid under up to 24kg payloads by combining Privileged Motion Guidance from noisy VR data with a Windowed Payload Curriculum.
-
ComplexMimic: Human-Scene Interaction Imitation in Complex 3D Environments
ComplexMimic applies a dual-flow imitation and interaction expert strategy plus difficulty-aware distillation to enable HSI mimicry in complex scenes and reports outperformance on three benchmarks.
-
Procedural Memory Distillation: Online Reflection for Self-Improving Language Models
PMD extracts and distills cross-episode procedural knowledge from RL rollouts into LLM policies at three abstraction levels, yielding 3.8-13.6% gains over SDPO on SCIKNOWEVAL and LIVECODEBENCH via co-evolution.
-
Language-Critique Imitation Learning from Suboptimal Demonstrations
Language-critique imitation learning constructs natural language labels from suboptimal demos to train LC-BC and LC-DP policies via a direct language loss, with a theoretical bound on expert gap and empirical gains on navigation, manipulation, and gameplay tasks.
-
Staleness-Learning Rate Scaling Laws for Asynchronous RLHF
Stale rollouts introduce O(S * eta) surrogate-gradient bias in async GRPO, yielding stability condition eta << min{R_batch / (S * G_upd), R_crit / (T * G_upd)} under smoothness assumptions.
-
Learning Gait-Aware Quadruped Locomotion with Temporal Logic Specifications
Framework using parameterized Signal Temporal Logic specifications to shape rewards for PPO-based RL, yielding tighter velocity tracking and more stable training than hand-crafted rewards on Barkour quadruped in MuJoCo simulation.
-
Stage-Transition Dense Reward Modeling for Reinforcement Learning
STDR infers stage structure from expert videos to supply stage-transition and within-stage progress rewards, improving RL sample efficiency on 14 manipulation tasks.
-
What Probing Reveals about Autonomous Driving: Linking Internal Prediction Errors to Ego Planning
Autonomous driving policies with strong closed-loop performance frequently lack timely internal predictions of surrounding vehicles during near-collision events, and causal correction of prediction errors leads to improved ego planning.
-
Sampling-Based Coordination-Informed Multi-Objective Multi-Robot Reinforcement Learning
CIMORL framework with sampling variants (TS and MPPI) uses privileged training for decentralized multi-objective multi-robot RL, reporting 21.2% hypervolume gain over baselines in cooperative and adversarial tests.
-
Experience Augmented Policy Optimization for LLM Reasoning
EAPO reuses prior RL policy experience adaptively at decision points in LLM rollouts with adapted importance sampling and reports gains over prior RLVR methods on math benchmarks.
-
FlowAWR: Online Adaptive Flow Reinforcement via Advantage-Weighted Rectification
FlowAWR derives an advantage-weighted rectification for optimal velocity fields in flow models, claiming 2-5x faster convergence than DiffusionNFT on SD3.5-Medium.
-
REAR: Test-time Preference Realignment through Reward Decomposition
REAR decomposes the reward into question and preference components, rescales their balance, and expresses the result as a linear combination of token log-probabilities for efficient integration with best-of-N and tree search.
-
Toward an Energy-Optimized Operation of Data Centers Located in Wind Farms Using Reinforcement Learning
Reinforcement learning with imitation learning and reward shaping improves online workload shifting in a one-turbine one-data-center simulation but remains below an offline optimizer that sees the full day.
-
AlgoSkill: Learning to Design Algorithms by Scheduling Human-Like Skills
AlgoSkill improves LLM algorithm design on programming benchmarks by framing it as verification-guided scheduling over a typed skill library with MCTS, outperforming direct generation and self-refinement.
-
RoamFlow: Reinforcement-Aligned One-Step Action MeanFlow Policy for Image-Goal Navigation
RoamFlow applies MeanFlow to predict average velocity fields for one-step action policies in image-goal navigation, trained via expert imitation followed by RL refinement.
-
Learned Coordination Conventions in Cooperative MARL: Measuring the Translation Gap Between Theory-Informed Roles and Learned Routing
Empirical diagnostic framework measuring alignment between theory-informed roles and learned routing conventions in cooperative MARL via attention models and attribution analysis.
-
UCOB: Learning to Utilize and Evolve Agentic Skills via Credit-Aware On-Policy Bidirectional Self-Distillation
UCOB improves agentic RL by using return-to-go comparisons between skill-conditioned and no-skill prompts as local teachers for bidirectional self-distillation and skill memory updates.
-
EntroRouter: Learning Efficient Model Routing via Entropy Regulation
EntroRouter applies entropy regulation in a single-round routing framework to decouple reasoning from routing, retaining 98.3% of top expert accuracy at 48.25% lower compute cost.
-
Behavior Uncloning: Distilling Mode Redirection into Policy Weights without Inference-Time Steering
MoRE improves robot policy success rates by 44 percentage points by distilling mode redirection into weights, matching filtered retraining performance without inference overhead.
-
GPC: Large-Scale Generative Pretraining for Transferable Motor Control
GPC learns a motion vocabulary via Finite Scalar Quantization and end-to-end RL, then trains an autoregressive transformer for next-token control generation, achieving 99.98% motion reproduction success with emergent robustness.
-
TrafficAlign: Aligning Large Language Models for Traffic Scenario Generation
TrafficAlign synthesizes scenarios from videos to align LLMs, producing tests that detect 10.8% more collisions than prior methods and enable 36.1% collision reduction after fine-tuning the driving models.
-
Regularized Reward-Punishment Reinforcement Learning
Introduces KCPR and its deep form klDMP that couples reward and punishment policies via learned priors, yielding improved safety and stability in grid-world and Gazebo navigation tasks over DQN, SQL and softDMP.
-
PPO-EAL: Exact Augmented Lagrangian Proximal Policy Optimization for Safe Robotic Control
PPO-EAL integrates exact augmented Lagrangian optimization into PPO for safe robotic control, with claimed theoretical guarantees and better empirical safety-performance tradeoffs on several robot benchmarks including sim-to-real gear assembly.
-
Position: RL Researchers Need to Distinguish Between Solving Simulators and Using Simulators as a Proxy
RL researchers need to separate 'solving simulators' from 'using simulators as proxies' to prevent misleading conclusions about algorithm performance.
-
Internalizing the Future: A Unified Agentic Training Paradigm for World Model Planning
A three-stage training pipeline internalizes world-model simulation and success estimation in LLM agents for improved planning on search and math tasks.
-
Deep Reinforcement Learning-Enhanced Event-Triggered Data-Driven Predictive Control for a 3D Cable-Driven Soft Robotic Arm
RL-ET-DeePC trains a model-free RL policy to trigger DeePC optimizations adaptively, cutting optimization frequency by up to 66% in simulation and 34% on hardware while preserving tracking accuracy for a soft robotic arm.
-
FlowTTS-GRPO: Online Reinforcement Learning with Multi-Objective Reward Optimization for Flow-Matching Based Text-to-Speech
FlowTTS-GRPO applies online RL with weighted multi-objective rewards to flow-matching TTS models via ODE-to-SDE conversion, reporting gains in speaker similarity and perceptual quality on CosyVoice 3.0 and F5-TTS.
-
Scalable Multi-Task Data Generation via Reinforcement Learning for Language-Conditioned Bimanual Dexterous Manipulation
An RL data generation pipeline with generalizable rewards and language annotations produces diverse synthetic datasets that improve multi-task policy generalization on three bimanual manipulation tasks.
-
Prefix-Guided On-Policy Distillation: Mining Golden Trajectories from Rollouts
PG-OPD uses early prefix overlap to selectively continue only high-compatibility rollouts in on-policy distillation, reporting up to 4.8 accuracy points gained and 2.46x less training time on math benchmarks.
-
Modularized Reinforcement Learning on LLMs: From MDP Creation to Exploration and Learning
Survey mapping RL techniques onto LLM training and highlighting gaps in value-based, off-policy, and bootstrapping methods.
-
Motion-Aware Reinforcement Learning For Object Localization
MARLNet adds a motion prior to observations and smoothness penalty to rewards in a PPO bounding-box agent, producing small gains (+0.011 on VOC, +0.007 on VisDrone) at IoU≥0.5 while exposing a reward-interference failure mode.
-
FAST: A Framework for Aligned Sampling and Training in Parallel Reinforcement Learning for Autonomous Driving
FAST uses Dynamic Parallel Sampling Alignment via virtual continuation and Scaled Mask-Padding Optimization to remove straggler bottlenecks in parallel RL, delivering 1.78x wall-clock speedup while preserving unbiasedness.