Evaluating Agents without Rewards

• URL: http://arxiv.org/abs/2012.11538v2
• Date: Tue, 9 Feb 2021 22:06:26 GMT
• Title: Evaluating Agents without Rewards
• Authors: Brendon Matusch, Jimmy Ba, Danijar Hafner
• Abstract summary: Competing objectives have been proposed for agents to learn without external supervision. We retrospectively compute potential objectives on pre-collected datasets of agent behavior. We find that all three intrinsic objectives correlate more strongly with a human behavior similarity metric than with task reward.
• Score: 33.17951971728784
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Reinforcement learning has enabled agents to solve challenging tasks in unknown environments. However, manually crafting reward functions can be time consuming, expensive, and error prone to human error. Competing objectives have been proposed for agents to learn without external supervision, but it has been unclear how well they reflect task rewards or human behavior. To accelerate the development of intrinsic objectives, we retrospectively compute potential objectives on pre-collected datasets of agent behavior, rather than optimizing them online, and compare them by analyzing their correlations. We study input entropy, information gain, and empowerment across seven agents, three Atari games, and the 3D game Minecraft. We find that all three intrinsic objectives correlate more strongly with a human behavior similarity metric than with task reward. Moreover, input entropy and information gain correlate more strongly with human similarity than task reward does, suggesting the use of intrinsic objectives for designing agents that behave similarly to human players.

Related papers

• Intrinsically-Motivated Humans and Agents in Open-World Exploration [50.00331050937369]
  We compare adults, children, and AI agents in a complex open-ended environment. We find that only Entropy and Empowerment are consistently positively correlated with human exploration progress.
  arXiv  Detail & Related papers  (2025-03-31T00:09:00Z)
• Generating Causal Explanations of Vehicular Agent Behavioural Interactions with Learnt Reward Profiles [13.450023647228843]
  We learn a weighting of reward metrics for agents such that explanations for agent interactions can be causally inferred. We validate our approach quantitatively and qualitatively across three real-world driving datasets.
  arXiv  Detail & Related papers  (2025-03-18T01:53:59Z)
• Enhancing Human Experience in Human-Agent Collaboration: A Human-Centered Modeling Approach Based on Positive Human Gain [18.968232976619912]
  We propose a "human-centered" modeling scheme for collaborative AI agents. We expect that agents should learn to enhance the extent to which humans achieve these goals while maintaining agents' original abilities. We evaluate the RLHG agent in the popular Multi-player Online Battle Arena (MOBA) game, Honor of Kings.
  arXiv  Detail & Related papers  (2024-01-28T05:05:57Z)
• Information Design in Multi-Agent Reinforcement Learning [61.140924904755266]
  Reinforcement learning (RL) is inspired by the way human infants and animals learn from the environment. Research in computational economics distills two ways to influence others directly: by providing tangible goods (mechanism design) and by providing information (information design)
  arXiv  Detail & Related papers  (2023-05-08T07:52:15Z)
• Improving Multimodal Interactive Agents with Reinforcement Learning from Human Feedback [16.268581985382433]
  An important goal in artificial intelligence is to create agents that can both interact naturally with humans and learn from their feedback. Here we demonstrate how to use reinforcement learning from human feedback to improve upon simulated, embodied agents.
  arXiv  Detail & Related papers  (2022-11-21T16:00:31Z)
• Human-Like Navigation Behavior: A Statistical Evaluation Framework [0.0]
  We build a non-parametric two-sample hypothesis test designed to compare the behaviors of artificial agents to those of human players. We show that the resulting $p$-value not only aligns with anonymous human judgment of human-like behavior, but also that it can be used as a measure of similarity.
  arXiv  Detail & Related papers  (2022-03-10T01:07:34Z)
• PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training [94.87393610927812]
  We present an off-policy, interactive reinforcement learning algorithm that capitalizes on the strengths of both feedback and off-policy learning. We demonstrate that our approach is capable of learning tasks of higher complexity than previously considered by human-in-the-loop methods.
  arXiv  Detail & Related papers  (2021-06-09T14:10:50Z)
• Learning Human Rewards by Inferring Their Latent Intelligence Levels in Multi-Agent Games: A Theory-of-Mind Approach with Application to Driving Data [18.750834997334664]
  We argue that humans are bounded rational and have different intelligence levels when reasoning about others' decision-making process. We propose a new multi-agent Inverse Reinforcement Learning framework that reasons about humans' latent intelligence levels during learning.
  arXiv  Detail & Related papers  (2021-03-07T07:48:31Z)
• Reinforcement Learning for Sparse-Reward Object-Interaction Tasks in a First-person Simulated 3D Environment [73.9469267445146]
  First-person object-interaction tasks in high-fidelity, 3D, simulated environments such as the AI2Thor pose significant sample-efficiency challenges for reinforcement learning agents. We show that one can learn object-interaction tasks from scratch without supervision by learning an attentive object-model as an auxiliary task.
  arXiv  Detail & Related papers  (2020-10-28T19:27:26Z)
• Action and Perception as Divergence Minimization [43.75550755678525]
  Action Perception Divergence is an approach for categorizing the space of possible objective functions for embodied agents. We show a spectrum that reaches from narrow to general objectives. These agents use perception to align their beliefs with the world and use actions to align the world with their beliefs.
  arXiv  Detail & Related papers  (2020-09-03T16:52:46Z)
• Learning to Incentivize Other Learning Agents [73.03133692589532]
  We show how to equip RL agents with the ability to give rewards directly to other agents, using a learned incentive function. Such agents significantly outperform standard RL and opponent-shaping agents in challenging general-sum Markov games. Our work points toward more opportunities and challenges along the path to ensure the common good in a multi-agent future.
  arXiv  Detail & Related papers  (2020-06-10T20:12:38Z)
• Intrinsic Motivation for Encouraging Synergistic Behavior [55.10275467562764]
  We study the role of intrinsic motivation as an exploration bias for reinforcement learning in sparse-reward synergistic tasks. Our key idea is that a good guiding principle for intrinsic motivation in synergistic tasks is to take actions which affect the world in ways that would not be achieved if the agents were acting on their own.
  arXiv  Detail & Related papers  (2020-02-12T19:34:51Z)

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.