ASPECT:Analogical Semantic Policy Execution via Language Conditioned Transfer

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Palattuparambil, Ajsal Shereef, Karimpanal, Thommen George, Rana, Santu
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915929386909696
author Palattuparambil, Ajsal Shereef
Karimpanal, Thommen George
Rana, Santu
author_facet Palattuparambil, Ajsal Shereef
Karimpanal, Thommen George
Rana, Santu
contents Reinforcement Learning (RL) agents often struggle to generalize knowledge to new tasks, even those structurally similar to ones they have mastered. Although recent approaches have attempted to mitigate this issue via zero-shot transfer, they are often constrained by predefined, discrete class systems, limiting their adaptability to novel or compositional task variations. We propose a significantly more generalized approach, replacing discrete latent variables with natural language conditioning via a text-conditioned Variational Autoencoder (VAE). Our core innovation utilizes a Large Language Model (LLM) as a dynamic \textit{semantic operator} at test time. Rather than relying on rigid rules, our agent queries the LLM to semantically remap the description of the current observation to align with the source task. This source-aligned caption conditions the VAE to generate an imagined state compatible with the agent's original training, enabling direct policy reuse. By harnessing the flexible reasoning capabilities of LLMs, our approach achieves zero-shot transfer across a broad spectrum of complex and truly novel analogous tasks, moving beyond the limitations of fixed category mappings. Code and videos are available \href{https://anonymous.4open.science/r/ASPECT-85C3/}{here}.
format Preprint
id arxiv_https___arxiv_org_abs_2604_08355
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle ASPECT:Analogical Semantic Policy Execution via Language Conditioned Transfer
Palattuparambil, Ajsal Shereef
Karimpanal, Thommen George
Rana, Santu
Artificial Intelligence
Reinforcement Learning (RL) agents often struggle to generalize knowledge to new tasks, even those structurally similar to ones they have mastered. Although recent approaches have attempted to mitigate this issue via zero-shot transfer, they are often constrained by predefined, discrete class systems, limiting their adaptability to novel or compositional task variations. We propose a significantly more generalized approach, replacing discrete latent variables with natural language conditioning via a text-conditioned Variational Autoencoder (VAE). Our core innovation utilizes a Large Language Model (LLM) as a dynamic \textit{semantic operator} at test time. Rather than relying on rigid rules, our agent queries the LLM to semantically remap the description of the current observation to align with the source task. This source-aligned caption conditions the VAE to generate an imagined state compatible with the agent's original training, enabling direct policy reuse. By harnessing the flexible reasoning capabilities of LLMs, our approach achieves zero-shot transfer across a broad spectrum of complex and truly novel analogous tasks, moving beyond the limitations of fixed category mappings. Code and videos are available \href{https://anonymous.4open.science/r/ASPECT-85C3/}{here}.
title ASPECT:Analogical Semantic Policy Execution via Language Conditioned Transfer
topic Artificial Intelligence
url https://arxiv.org/abs/2604.08355