Skip to main content
GameDev.net gamedev.net

Reinforcement Learning News

The latest Reinforcement Learning coverage curated for game developers.

ViCo: Visual-oriented Coding with Self-Reflection for Chart Replication

ViCo is a new training framework aimed at making AI-generated charts look and read more like human-made figures. It combines self-reflection, multi-step reinforcement learning, and automated visual checks to improve …

Research arXiv cs.GR · 1 week, 5 days ago
Learning Realistic Athletic Sprinting Without Demonstrations

A new muscle-driven simulation pipeline can generate realistic sprinting and drill motions without motion-capture demonstrations. Running at up to 1000x real time on a single GPU, it trains control policies …

Research arXiv cs.GR · 2 weeks, 3 days ago
Reflex-Informed Neuromuscular Reinforcement Learning for Muscle-Driven Locomotion

A new muscle-driven locomotion system uses a fixed reflex controller plus reinforcement learning to tune just four biomechanical parameters. The result is more plausible gait, better symmetry, and stronger robustness …

Research arXiv cs.GR · 2 weeks, 3 days ago
InstantMimic: A High Performance System for Learning Physics-based Skills in Seconds

InstantMimic pushes physics-based character training into the seconds range by moving the entire RL loop onto the GPU. The system targets imitation-driven motion control, cutting out CPU bottlenecks and fragmented …

Research arXiv cs.GR · 2 weeks, 4 days ago
Lambda-Hold Control: Human-Like Movement Emerges from a Minimal Task Reward in Predictive Musculoskeletal Simulation

A new musculoskeletal RL controller can produce human-like sprinting with only a minimal task reward and about an hour of training. The ?bb-hold approach cuts the action space down to …

Research arXiv cs.GR · 1 month, 1 week ago
RealMat: Realistic Materials with Diffusion and Reinforcement Learning

RealMat combines Stable Diffusion XL with reinforcement learning to generate more believable material maps for 3D authoring. The pipeline starts from synthetic 2×2 material grids, then pushes the model toward …

Research arXiv cs.GR · 1 month, 2 weeks ago
RL-Lock: Reinforcement Learning for Generating Interlocking Assemblies

RL-Lock brings reinforcement learning to interlocking assembly generation, turning voxel decomposition into a sequential decision problem. The system uses structured action chunking plus MCTS-guided policy-value learning to search huge combinatorial …

Research arXiv cs.GR · 1 month, 3 weeks ago
ThinkBLOX: 3D Indoor Scene Generation with Progressive Reasoning

ThinkBLOX is a new VLM-driven pipeline for generating 3D indoor scenes through progressive reasoning instead of one-shot layout planning. It iteratively places and refines objects, aiming to reduce the awkward …

Research arXiv cs.GR · 2 months, 1 week ago
GPC: Large-Scale Generative Pretraining for Transferable Motor Control

A SIGGRAPH 2026 paper reports a 99.98% motion-reproduction success rate using a tokenized, GPT-style controller trained on large motion datasets. For animation teams, the interesting part is the shift from …

Research arXiv cs.GR · 2 months, 4 weeks ago
Latent Space Reinforcement Learning for Inverse Material Estimation in Food Fracture Simulation

Researchers trained a goal-conditioned PPO policy to estimate food material parameters from fracture behavior in a single forward pass, then used CMA-ES to refine the result. On orange peeling, the …

Research arXiv cs.GR · 3 months, 1 week ago
HIL: Hybrid Imitation Learning of Diverse Parkour Skills from Videos

This paper shows a way to train one controller that can both mimic reference parkour motion and still adapt when the environment changes. For teams working on character movement, the …

Research arXiv cs.GR · 3 months, 1 week ago
TacCoRL: Integrating Tactile Feedback into VLA via Simulation

TacCoRL reports a 72.5% average success rate on four contact-heavy bimanual tasks by adding tactile feedback to VLA policies and training in simulation. The practical takeaway is that the model …

Research arXiv cs.GR · 3 months, 2 weeks ago
SCRIPT: Scalable Diffusion Policy with Multi-stage Training for Language-driven Physics-Based Humanoid Control

SCRIPT pushes language-driven humanoid control toward more stable, physically plausible motion. The system combines a diffusion policy with multi-stage training, then adds reinforcement learning to improve instruction following and motion …

Research arXiv cs.GR · 4 months ago
NaP-Control: Navigating Diffusion Prior for Versatile and Fast Character Control

NaP-Control uses reinforcement learning to steer a diffusion policy’s latent noise, aiming for whole-body character control that is both fast and robust. The approach skips iterative test-time guidance, which can …

Research arXiv cs.GR · 4 months, 1 week ago
Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration

A new arXiv paper describes COSMO-Agent, a tool-augmented RL setup that lets an LLM drive a closed-loop CAD-to-simulation workflow. The interesting bit for game teams is the orchestration pattern: external …

Research arXiv cs.GR · 4 months, 1 week ago
Announcing AMD Schola v2.1: state trees, scale, and a richer training stack

AMD Schola v2.1 adds StateTree support and a more scalable training pipeline, which makes it more practical for teams using Unreal Engine to train and deploy agent behavior. The big …

Official AMD GPUOpen · 4 months, 1 week ago
ReActor: Reinforcement Learning for Physics-Aware Motion Retargeting

A new framework for motion retargeting using reinforcement learning addresses common issues like foot sliding and self-collisions. This innovation is particularly relevant for programmers and animators working with robotics and …

Research arXiv cs.GR · 4 months, 2 weeks ago
Training and Agentic Inference Strategies for LLM-based Manim Animation Generation

A new study shows LLMs can generate Manim animations more reliably when training and inference are treated as separate problems. The strongest setup paired supervised fine-tuning with GRPO and a …

Research arXiv cs.GR · 5 months, 1 week ago
COSMO-Agent: Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration

COSMO-Agent introduces a novel tool-augmented reinforcement learning framework that bridges the CAD-CAE semantic gap, enhancing iterative design processes. This advancement is particularly relevant for developers involved in optimization and simulation, …

Research arXiv cs.GR · 5 months, 2 weeks ago
Teaching an Agent to Sketch One Part at a Time

A new method for training agents to create vector sketches part by part is making waves in AI and graphics. This approach leverages a unique dataset and reinforcement learning, offering …

Research arXiv cs.GR · 6 months ago
Page 1 of 2 Next