mlsanityData to 7 Oct 2026

From Path Signatures to Sequential Modeling: Incremental Signature Contributions for Offline RL

arXiv 2602.11805 ↗First posted 12 Feb 2026cs.LGcited by 0

On the map: Offline RL and imitation learning from data → Decision transformers and sequence modeling

Abstract (arXiv)

Path signatures embed trajectories into tensor algebra and constitute a universal, non-parametric representation of paths; however, in the standard form, they collapse temporal structure into a single global object, which limits their suitability for decision-making problems that require step-wise reactivity. We propose the Incremental Signature Contribution (ISC) method, which decomposes truncated path signatures into a temporally ordered sequence of elements in the tensor-algebra space, corresponding to incremental contributions induced by last path increments. This reconstruction preserves the algebraic structure and expressivity of signatures, while making their internal temporal evolution explicit, enabling processing signature-based representations via sequential modeling approaches. In contrast to full signatures, ISC is inherently sensitive to instantaneous trajectory updates, which is critical for sensitive and stability-requiring control dynamics. Building on this representation, we introduce ISC-Transformer (ISCT), an offline reinforcement learning model that integrates ISC into a standard Transformer architecture without further architectural modification. We evaluate ISCT on HalfCheetah, Walker2d, Hopper, and Maze2d, including settings with delayed rewards and downgraded datasets. The results demonstrate that ISC method provides a theoretically grounded and practically effective alternative to path processing for temporally sensitive control tasks.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Decision transformers and sequence modeling, one of 13 directions in Offline RL and imitation learning from data. Builds on 18 of its 25 references that are in the map. First posted on arXiv 12 Feb 2026.

Builds on

18 of 25 references.

Closest ancestors · 5

Same direction, newest first

  1. 2022Hierarchical Decision Transformer21 Sep
  2. 2022Online Decision Transformer11 Feb
  3. 2021Generalized Decision Transformer for Offline Hindsight Information Matching19 Nov
  4. 2021Offline Reinforcement Learning as One Big Sequence Modeling Problem3 Jun
  5. 2021Decision Transformer: Reinforcement Learning via Sequence Modeling2 Jun

Other references · 13

  1. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  2. 2020Conservative Q-Learning for Offline Reinforcement Learningcited by 1,777
  3. 2020D4RL: Datasets for Deep Data-Driven Reinforcement Learningcited by 1,406
  4. 2021Offline Reinforcement Learning with Implicit Q-Learningcited by 1,245
  5. 2021A Minimalist Approach to Offline Reinforcement Learningcited by 882
  6. 2022Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learningcited by 543
  7. 2020Neural Controlled Differential Equations for Irregular Time Seriescited by 501
  8. 2020Neural Rough Differential Equations for Long Time Seriescited by 140
  9. 2021Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPscited by 115
  10. 2024Log Neural Controlled Differential Equations: The Lie Brackets Make a Differencecited by 34
  11. 2013Sparse arrays of signatures for online character recognitioncited by 17
  12. 2023Learnable Path in Neural Controlled Differential Equationscited by 8
1 more
  1. 2023SigFormer: Signature Transformers for Deep Hedgingcited by 7

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026MAGE: Multi-scale Autoregressive Generation for Offline Reinforcement Learning
  2. 2023Multi-Objective Decision Transformers for Offline Reinforcement Learning
  3. 2026PathBridger: Subgoal Bridges for Offline Goal-Conditioned Reinforcement Learning
  4. 2024M3PC: Test-time Model Predictive Control for Pretrained Masked Trajectory Model
  5. 2025SAC Flow: Sample-Efficient Reinforcement Learning of Flow-Based Policies via Velocity-Reparameterized Sequential Modeling
  6. 2026Hitting Time Isomorphism for Multi-Stage Planning with Foundation Policies
  7. 2026Preserve Support, Not Correspondence: Dynamic Routing for Offline Reinforcement Learning
  8. 2024Q-value Regularized Transformer for Offline Reinforcement Learning