Proposes using logical specifications for multi-objective reinforcement learning to improve agent behavior.
problem Difficulties in controlling reinforcement learning agents and the need for better generalization.
method Uses propositional logic to specify the importance of multiple objectives, encoding these specifications using a recurrent neural network.
result MORL agents parameterized by logical specifications can generalize to novel combinations of objectives and achieve comparable performance.
New approach categorizes objective functions for embodied agents.
problem Understanding how objectives relate to each other and discovering new objectives.
method Introducing Action Perception Divergence (APD) to categorize objective functions.
result Introduces a spectrum of objectives from narrow to general, explaining various unsupervised objectives.
Agents use object-oriented reasoning to solve problems more quickly.
problem Finding solutions to problems more efficiently.
method Hierarchical controller directs low-level agent to simulate alternate states of the world.
result Achieves similar reward levels as non-hierarchical agents but with better data efficiency.
Algorithm for agents to agree on a single objective in multi-task networks.
problem Decentralized decision-making in multi-task networks with multiple objectives.
method Distributed decision-making algorithm for agents observing different models.
result Agents reach agreement on which model to track for network performance enhancement.
Study how financial market participants have different objectives at various time scales.
problem Understanding how financial market participants have different objectives at various time scales.
method Use Inverse Reinforcement Learning to compute the effective reward function for the aggregate agent class at each scale.
result Identify differences in reward functions for feature vectors across different scales, indicating different objectives of market participants.
Deep Sets improve reinforcement learning agent's object-centered navigation and generalization.
problem Improving reinforcement learning agents' ability to generalize to unseen objects and goals.
method Combining object-wise permutation invariant networks (Deep Sets) and gated-attention mechanisms.
result Agent demonstrates strong generalization to out-of-distribution goals in a procedurally-generated 2D world.
We consider a set of learning agents in a collaborative peer-to-peer network, where each agent learns a personalized model according to its own learning objective. The question addressed in this paper is: how can agents improve upon their locally trained model by communicating with other agents that have similar object…
Agents learn to play by predicting their own errors and exploring novel interactions.
problem Creating autonomous agents that can learn and explore in complex, unstructured environments.
method A neural network with a world-model and self-model that learns to predict and challenge its own predictions.
result The agent generates complex behaviors like ego-motion prediction and object gathering.
Agent finds hidden objects in DAGs, stopping and restarting.
problem Finding hidden objects in a directed acyclic graph.
method Sequential search-and-stop strategy, learning from unknown distributions.
result Near-optimal strategy for unknown distributions, maximizing object collection.
The paper proposes a method to infer multi-objective rewards from preferences.
problem Modeling preferences based on multiple, often competing objectives.
method Modeling priorities lexicographically and inferring multi-objective rewards from observed preferences.
result Lexicographically-ordered rewards provide a better understanding of preferences and improve policies.
New algorithm minimizes FE objectives for synthetic AIF agents.
problem Minimizing Free Energy for synthetic active inference agents.
method Variational message passing on Constrained FFGs.
result Message passing algorithms induce epistemic behavior in AIF agents.
Paper identifies invariant structures for object perception.
problem Understanding how agents perceive objects in dynamic environments.
method Sensorimotor Contingencies Theory inspired unsupervised predictive model.
result Agents can identify invariant structures in sensorimotor experiences.
Method estimates forces from agent trajectories to infer static obstacles.
problem Estimating forces from agent trajectories to infer static obstacles.
method Artificial neural networks to estimate non-parametric velocity fields.
result Incrementally learns velocity fields due to static objects.
We analyze systems of agents sharing light-tailed risky claims issued by different financial objects. Assuming exponentially distributed claims, we obtain that both agents' and system's losses follow generalized exponential mixture distributions. We show that this leads to qualitatively different results on individual …
New architecture tracks objects in cluttered scenes without supervision.
problem Learning to track objects in large scenes with many objects.
method Spatially invariant computations and representations.
result Outperforms competing methods in tracking objects in cluttered scenes.
Neural network learns to explore and predict actions in a simulated environment.
problem Replicate infants' ability to generate structured behaviors in unstructured environments.
method Curiosity-driven intrinsic motivation, adversarial learning, world model prediction.
result Self-supervised emergence of complex behaviors including object gathering.
Agent learns object segmentation through self-supervised interactions.
problem Object segmentation in visual observations.
method Self-supervised learning through interaction, robust set loss.
result Model generalizes to novel objects and backgrounds.
New algorithm improves performance in nontransitive games.
problem Nontransitive games lack a clear winner.
method Geometric framework for agent objectives, PSRO_rN algorithm.
result PSRO_rN consistently outperforms alternatives in nontransitive games.
Agents learn sophisticated tool use and coordination in hide-and-seek.
problem Training agents to perform complex tasks in multi-agent environments.
method Multi-agent competition with hide-and-seek as a simple objective.
result Agents develop emergent strategies requiring sophisticated tool use and coordination.
MERL uses evolutionary and gradient-based methods to optimize sparse team-based and dense agent-specific rewards in multiagent coordination.
problem Training multiagent reinforcement learning policies on sparse team-based rewards is difficult and relying solely on agent-specific rewards is sub-optimal.
method MERL employs a split-level training platform with an evolutionary algorithm and a gradient-based optimizer, transferring skills between the two processes.
result MERL significantly outperforms state-of-the-art methods on coordination benchmarks.
FEN model learns fairness and efficiency in multi-agent systems.
problem Fairness in multi-agent systems for stability and productivity.
method Hierarchical reinforcement learning model with fair-efficient reward and sub-policies.
result FEN model easily learns both fairness and efficiency in multi-agent scenarios.
Action guidance helps agents learn true objectives in games with sparse rewards.
problem Training agents in games with sparse rewards requires significant exploration.
method Action guidance, a novel technique that combines exploration with reward shaping.
result Action guidance enables agents to optimize true objectives efficiently.
CLIC learns to control objects in non-rewarding environments by imitating a mentor.
problem Learning in non-rewarding, multi-object environments with non-observable intentions.
method Curriculum Learning and Imitation (CLIC) for Control.
result CLIC effectively learns and imitates a mentor's interactions to control objects.
A decentralized approach for agents to learn and optimize collectively.
problem Challenges in coordinating non-cooperative agents to solve complex sequential decision problems.
method Designing a learning environment where agents learn by trading and optimizing local objectives, leading to a Nash equilibrium.
result Decentralized reinforcement learning algorithms that can handle various decision-making scenarios.
Distributed algorithm converges to global optima with Gaussian noise.
problem Nonconvex optimization in network-based multi-agent settings.
method Consensus+innovations type algorithm with annealing-type approach and Gaussian noise.
result Convergence to the set of global optima of the sum function.
We build deep RL agents that execute declarative programs expressed in formal language. The agents learn to ground the terms in this language in their environment, and can generalize their behavior at test time to execute new programs that refer to objects that were not referenced during training. The agents develop di…
Study designs incentives for adapting multi-agent systems without knowing their learning dynamics.
problem Designing incentives for an adapting population in multi-agent systems without prior knowledge of their learning dynamics.
method Introduces a model-based non-episodic Reinforcement Learning (RL) formulation for steering Markovian agents towards desired policies, focusing on history-dependent strategies to handle model uncertainty.
result Identifies conditions for the existence of steering strategies to guide agents to desired policies and provides empirical algorithms to approximately solve the objective.
Deep learning agent improves pedestrian navigation in urban environments.
problem Autonomous driving among pedestrians in urban areas.
method Multi-objective deep reinforcement learning using a deep Q-learning variant.
result The multi-objective DQN agent outperforms single-objective DQN in various environments.
A novel RL objective and prioritization framework improve performance and sample-efficiency in multi-goal tasks.
problem Learning diverse goals in multi-goal reinforcement learning.
method Maximum entropy regularization for objective and prioritization framework.
result Promising improvements in performance and sample-efficiency on multi-goal robotic tasks.
Efficiently learns uniform state distributions in unknown MDPs.
problem Learning uniform state distributions in unknown MDPs without rewards.
method Uses conditional gradient method with approximate MDP solver.
result Provable efficiency in sample and computational complexities.
Agent learns from an expert, adapting to constraints in concept learning.
problem Insufficient query selection in active learning for realistic human domains.
method Imitation learning to reason about both internal goals and external constraints.
result Agent outperforms other active learners under most constrained conditions.
We consider the problem of decentralized clustering and estimation over multi-task networks, where agents infer and track different models of interest. The agents do not know beforehand which model is generating their own data. They also do not know which agents in their neighborhood belong to the same cluster. We prop…
Enhances RL agents with predictive internal representations.
problem Improving model-free reinforcement learning agents.
method Introduces Deep InfoMax (DIM) objective to train predictive internal representations.
result Successfully learned predictive representations in synthetic settings.
Distributed processing over networks relies on in-network processing and cooperation among neighboring agents. Cooperation is beneficial when agents share a common objective. However, in many applications agents may belong to different clusters that pursue different objectives. Then, indiscriminate cooperation will lea…
Many practical environments contain catastrophic states that an optimal agent would visit infrequently or never. Even on toy problems, Deep Reinforcement Learning (DRL) agents tend to periodically revisit these states upon forgetting their existence under a new policy. We introduce intrinsic fear (IF), a learned reward…
This essay discusses the advantages of a probabilistic agent-based approach to questions in theoretical economics, from the nature of economic agents, to the nature of the equilibria supported by their interactions. One idea we propose is that "agents" are meta-individual, hierarchically structured objects, that includ…
Generative thermal design learns optimal shapes using multi-agent reinforcement learning.
problem Complex thermal design challenges due to convection-diffusion equation and boundary interactions.
method Cooperative multi-agent deep reinforcement learning with continuous geometric representation.
result Framework learns optimal design strategies without shape derivation or differentiable objectives.
Dreamer 4 learns Minecraft tasks from videos alone.
problem Accurately predicting object interactions in complex environments.
method Reinforcement learning inside a fast, accurate world model.
result Dreamer 4 outperforms previous models in Minecraft, learning from only offline data.
We present a distributed proximal-gradient method for optimizing the average of convex functions, each of which is the private local objective of an agent in a network with time-varying topology. The local objectives have distinct differentiable components, but they share a common nondifferentiable component, which has…
Adversaries can manipulate cooperative MARL networks.
problem Fragility of consensus-based MARL algorithms to adversarial attacks.
method Study of adversarial attacks on consensus-based MARL networks.
result Consensus-based MARL algorithms are vulnerable to adversarial manipulation.
We develop a model to study the role of rationality in economics and biology. The model's agents differ continuously in their ability to make rational choices. The agents' objective is to ensure their individual survival over time or, equivalently, to maximize profits. In equilibrium, however, rational agents who maxim…
Deep RL agent outperforms humans in constructing towers using relational reasoning.
problem Current deep learning systems struggle with constructing and modifying complex systems.
method Introduced a deep reinforcement learning agent with object- and relation-centric scene and policy representations.
result Structured representations allow the agent to outperform humans and naive approaches.
A new reinforcement learning method uses mutual information to encourage agents to control their environment.
problem Learning from internal drives instead of external rewards.
method Formulate an intrinsic objective as mutual information between goal states and controllable states, derive a surrogate objective for efficient optimization.
result Demonstrated the efficacy of the approach in robotic tasks.
Paper studies competitive networks where teams aim to minimize their own objectives, adapting to each other's strategies.
problem Competitive networks where teams have conflicting objectives.
method Proposes diffusion learning algorithms for two classes of network games: zero-sum and non-zero-sum.
result Stability performance of proposed algorithms analyzed and demonstrated through experiments.
VNLA uses vision and language to guide agents in finding objects in indoor environments.
problem Guiding agents in finding objects in indoor environments via language.
method Developed I3L framework for imitation learning with indirect intervention.
result Significantly improved success rate of learning agents over baselines.
A multi-stage reinforcement learning method for object detection.
problem Efficiently detecting objects within images with high accuracy.
method Hierarchical tree-like region candidates, zoom and refinement stages, aspect ratio modification, multiple reward metrics.
result The multi-stage approach leads to more correct detections compared to single-stage methods.
Mobile agents classify images via reinforcement learning and consensus.
problem Image classification using multiple mobile agents.
method Proposed network architecture for local belief formation and feature extraction. Decentralized consensus protocol using reinforcement learning.
result Effectiveness of the proposed framework demonstrated on MNIST dataset.
A framework disentangles controllable objects from visual signals for improved RL.
problem Improving sample efficiency and game performance in vision-based RL.
method Action-conditioned video prediction to disentangle controllable objects.
result Improved sample efficiency and game performance in Atari games.