mlsanityData to 14 Oct 2026

Probabilistic Recurrent Intention Switching Model

arXiv 2605.26998 ↗First posted 26 May 2026cs.LG q-bio.NCcited by 0

On the map: Offline RL and imitation learning from data → Inverse reinforcement learning

Abstract (arXiv)

Inverse reinforcement learning (IRL) recovers reward functions from observed behavior, yet traditional methods assume a single stationary reward that cannot capture goal switching within an episode. Recent multi-intention IRL methods address this by segmenting trajectories, but model intention transitions as either a memoryless Markov chain or via manual state augmentation with a fixed history window. We propose the Probabilistic Recurrent Intention Switching Model (PRISM), which replaces both mechanisms with a lightweight recurrent network that maps observation history to a per-step intention distribution. We prove that the resulting EM objective decomposes exactly into independent per-intention reward subproblems, each solvable in closed form, yielding an O(nK) E-step with no variational approximation. We evaluate PRISM on a non-Markovian gridworld, a mouse labyrinth, and BridgeData~V2 robotic manipulation, the first large-scale robotic application of multi-intention IRL. Across all settings PRISM achieves the highest held-out log-likelihood while recovering nameable, temporally coherent intentions from unlabeled demonstrations, suggesting that discrete goal switching is present in both biological and artificial agents.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Inverse reinforcement learning, one of 13 directions in Offline RL and imitation learning from data. Builds on 14 of its 27 references that are in the map. First posted on arXiv 26 May 2026.

Builds on

14 of 27 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Inverse Reinforcement Learning with Switching Rewards and History Dependency for Characterizing Animal Behaviors22 Jan
  2. 2023Multi-intention Inverse Q-learning for Interpretable Behavior Representation23 Nov
  3. 2020Deep Inverse Q-learning with Constraints4 Aug
  4. 2018A Survey of Inverse Reinforcement Learning: Challenges, Methods and Progress18 Jun

Other references · 10

  1. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  2. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,061
  3. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,761
  4. 2016Generative Adversarial Imitation Learningcited by 1,597
  5. 2025SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Featurescited by 1,188
  6. 2023BridgeData V2: A Dataset for Robot Learning at Scalecited by 657
  7. 2019Learning Latent Plans from Playcited by 316
  8. 2018Programmatically Interpretable Reinforcement Learningcited by 193
  9. 2024Behavior Generation with Latent Actionscited by 160
  10. 2023Action-Quantized Offline Reinforcement Learning for Robotic Skill Learningcited by 36

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2022Basis for Intentions: Efficient Inverse Reinforcement Learning using Past Experience
  2. 2021Deep Adaptive Multi-Intention Inverse Reinforcement Learning
  3. 2019Meta-Inverse Reinforcement Learning with Probabilistic Context Variables
  4. 2019From Language to Goals: Inverse Reinforcement Learning for Vision-Based Instruction Following
  5. 2018Inverse Reinforcement Learning via Nonparametric Spatio-Temporal Subgoal Modeling
  6. 2025Inversely Learning Transferable Rewards via Abstracted States
  7. 2026Graph-Reweighted Behavior Cloning with Human Interventions for Precise Robotic Manipulation
  8. 2024Toward Information Theoretic Active Inverse Reinforcement Learning