Latent Dynamics · 2026-10-12
RoleWorld: Knowledge Transfer in World Modeling with Functional Role Abstractions
Boyuan Xiao, Pu Wang, Ji Feng, Yao-Xiang Ding
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
RoleWorld is a symbolic world modeling approach for knowledge transfer. A teacher LLM extracts reusable role prototypes and executable transition rules, plus a role-guided skill from source trajectories. In a target world without the teacher, the actor keeps role prototypes/rules and the skill fixed, refines object–role assignments and environment-specific parameter estimates online, and uses active verification when uncertainty matters. Experiments on MARS report robust generalization across environment variants with altered object behavior.
Why it matters
The paper targets a transfer bottleneck in symbolic program/rule world models: rules tied to environment-specific objects can fail when object functions change. RoleWorld instead factors shared interaction mechanisms (role prototypes + transition rules) from environment-specific instantiation (role assignments + parameter values), enabling adaptation via online refinement while keeping the induced transition library fixed.
Method
- Teacher LLM builds a role-based symbolic world model from source trajectories: reusable role prototypes and executable transition rules defined over role instances in a DSL.
- In target worlds, the actor runs without the teacher: role prototypes, transition rules, and an actor skill stay fixed; online inference refines object–role assignments and environment-specific role parameters from interaction feedback.
- The skill organizes role information for decisions and guides active verification: the actor selects informative interactions when role-parameter or role-assignment uncertainty could affect the current decision.
Limitation
The paper does not establish correctness for other role assignments, all reachable states, or complete world-state predictions (in the context of its environment-transition validation of a frozen reference prior).
Abstract (from arXiv)
World models enable large language model (LLM) agents to anticipate action effects and make informed decisions. For generalization, it is desirable for such world models to transfer efficiently to new environments with similar interaction structure and underlying dynamics. Recent approaches encode world dynamics as symbolic rules or executable programs, whose compositional structure in principle affords transfer, but still organize such knowledge around environment-specific objects. To address this limitation, we introduce RoleWorld, a symbolic world modeling approach which uses reusable prototypes and environment-specific parameter values as functional role abstractions to represent how objects behave in interactions. From interaction trajectories, a teacher LLM extracts role prototypes and executable transition rules defined over role instances, together with a skill that guides the actor LLM to use role information in decision making. In a target world, without the teacher, the actor uses this skill together with its current role assignments and parameter estimates, and actively tests a role instance hypothesis through interaction when its uncertainty affects the current decision. Feedback from each action refines role assignments and estimates parameter values online, with no LLM parameter updates and the transition rules held fixed. Experiments on MARS benchmark show that RoleWorld enables robust generalization across environments with significantly altered object behavior, suggesting that functional role abstractions can serve as an effective component for knowledge transfer in world modeling. Project Page: https://future-item.github.io/RoleWorld.
Related papers
- AnyGoal: On World Model Generalization Beyond Expert Rollouts
- Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks
- From Single-Actor Environments to Multi-Actor Event Worlds: Can LLMs Be Event World Models?
- What Matters for World Modeling in Vision-Language-Action Models?
- EvoBelief: Falsifiable World Models for Self-Improving LLM Agents
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First public release: 2026-10-09 · Indexed by Pub-AI: 2026-10-12 UTC
RoleWorld: Knowledge Transfer in World Modeling with Functional Role Abstractions
Preprint posted: 2026-10-09 · Original record ↗
- arxiv
- 2610.13172
Explicit source versions (1)
- v1 · 2026-10-09
- Preprint available
Source date: 2026-10-09 · First detected: 2026-10-25 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · atom:published (official metadata)
Abstract · arxiv
World models enable large language model (LLM) agents to anticipate action effects and make informed decisions. For generalization, it is desirable for such world models to transfer efficiently to new environments with similar interaction structure and underlying dynamics. Recent approaches encode world dynamics as symbolic rules or executable programs, whose compositional structure in principle affords transfer, but still organize such knowledge around environment-specific objects. To address this limitation, we introduce RoleWorld, a symbolic world modeling approach which uses reusable prototypes and environment-specific parameter values as functional role abstractions to represent how objects behave in interactions. From interaction trajectories, a teacher LLM extracts role prototypes and executable transition rules defined over role instances, together with a skill that guides the actor LLM to use role information in decision making. In a target world, without the teacher, the actor uses this skill together with its current role assignments and parameter estimates, and actively tests a role instance hypothesis through interaction when its uncertainty affects the current decision. Feedback from each action refines role assignments and estimates parameter values online, with no LLM parameter updates and the transition rules held fixed. Experiments on MARS benchmark show that RoleWorld enables robust generalization across environments with significantly altered object behavior, suggesting that functional role abstractions can serve as an effective component for knowledge transfer in world modeling. Project Page: https://future-item.github.io/RoleWorld.