Dream-RSI: Recursive Self-Improvement through Evolving Worlds

• URL: http://arxiv.org/abs/2609.14858v1
• Date: Mon, 14 Sep 2026 00:10:47 GMT
• Title: Dream-RSI: Recursive Self-Improvement through Evolving Worlds
• Abstract summary: textscDream-RSI is a framework for scalable self-improving exploration.<n>A lightweight orchestration layer makes exploration explicit and programmable.<n>By performing dreaming in the replay simulator constructed from historical discovery trees, textscDream-RSI secures immediate, low-cost off-policy feedback.
• Score: 61.888486670260306
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recursive self-improvement is becoming increasingly vital for autonomous AI agents, where progress hinges on discovering high-value solutions across complex domains. The driver of this process is effective exploration, however, managing and improving exploration strategies remains a major bottleneck. Current systems face a fundamental dilemma: fixed strategies fail to adapt as search spaces scale, while online policy optimization requires navigating vast meta-search spaces under delayed and expensive feedback over long-horizon rollouts. We introduce \textsc{Dream-RSI}, a framework for scalable and recursively self-improving exploration. A lightweight orchestration layer makes exploration explicit and programmable while leaving the underlying coding agent unchanged. Our key insight is that accumulated discovery history can serve as a replay simulator over the realized search space. By performing dreaming in the replay simulator constructed from historical discovery trees, \textsc{Dream-RSI} secures immediate, low-cost off-policy feedback to evaluate and refine exploration policies without invoking repetitive, expensive online evaluations. The improved policy is subsequently redeployed online to drive further discovery, continuously expanding the simulator pool in a self-improving loop. Across algorithm engineering, mathematical optimization, and GPU kernel engineering, \textsc{Dream-RSI} achieves competitive or improved discovery quality while substantially reducing discovery cost in several settings.

Related papers

• Maximum Entropy Behavior Exploration for Sim2Real Zero-Shot Reinforcement Learning [18.905272125661824]
  We study $textitonline$ zero-shot RL for quadrupedal control on real robotic systems.<n>We introduce FB-MEBE, an online zero-shot RL algorithm that combines an unsupervised behavior exploration strategy with a regularization critic.
  arXiv  Detail & Related papers  (2026-03-26T14:07:01Z)
• Decoupling Exploration and Policy Optimization: Uncertainty Guided Tree Search for Hard Exploration [12.531650952835493]
  In this paper, we propose a new paradigm that explicitly separates exploration from exploitation and bypasses RL during the exploration phase.<n>By removing the overhead of policy optimization, our approach explores an order of magnitude more efficiently than standard intrinsic motivation baselines on hard Atari benchmarks.<n>We demonstrate that the discovered trajectories can be distilled into deployable policies using existing supervised backward learning algorithms.
  arXiv  Detail & Related papers  (2026-03-23T17:56:52Z)
• PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution [64.15555230987222]
  PACEvolve is a framework designed to robustly govern the agent's context and search dynamics.<n>We demonstrate that PACEvolve provides a systematic path to consistent, long-horizon self-improvement.
  arXiv  Detail & Related papers  (2026-01-15T18:25:23Z)
• DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search [53.27052683356095]
  We present DeepSearch, a framework that integrates Monte Carlo Tree Search directly into RLVR training.<n>In contrast to existing methods that rely on tree search only at inference, DeepSearch embeds structured search into the training loop.<n>Our contributions include: (1) a global frontier selection strategy that prioritizes promising nodes across the search tree, (2) selection with entropy-based guidance that identifies confident paths for supervision, and (3) adaptive replay buffer training with solution caching for efficiency.
  arXiv  Detail & Related papers  (2025-09-29T20:00:29Z)
• DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RL [60.47878242100153]
  We present DeepDive to advance deep search agents.<n>We propose a strategy to automatically synthesize complex, difficult, and hard-to-find questions from open knowledge graphs.<n>We apply end-to-end multi-turn reinforcement learning to enhance LLMs' long-horizon reasoning with deep search.
  arXiv  Detail & Related papers  (2025-09-12T17:52:35Z)
• SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis [94.33978856270268]
  Retrieval-augmented generation (RAG) systems have advanced large language models (LLMs) in complex deep search scenarios.<n>Existing approaches face critical limitations that lack high-quality training trajectories and suffer from distributional mismatches.<n>This paper introduces SimpleDeepSearcher, a framework that bridges the gap through strategic data engineering rather than complex training paradigms.
  arXiv  Detail & Related papers  (2025-05-22T16:05:02Z)
• Knowing You Don't Know: Learning When to Continue Search in Multi-round RAG through Self-Practicing [4.874077691069634]
  Retrieval Augmented Generation (RAG) has shown strong capability in enhancing language models' knowledge and reducing AI generative hallucinations.<n>Current multi-round RAG systems may continue searching even when enough information has already been retrieved.<n>This paper introduces a new framework, SIM-RAG, to explicitly enhance RAG systems' self-awareness and multi-round retrieval capabilities.
  arXiv  Detail & Related papers  (2025-05-05T17:39:35Z)
• Reannealing of Decaying Exploration Based On Heuristic Measure in Deep Q-Network [82.20059754270302]
  We propose an algorithm based on the idea of reannealing, that aims at encouraging exploration only when it is needed. We perform an illustrative case study showing that it has potential to both accelerate training and obtain a better policy.
  arXiv  Detail & Related papers  (2020-09-29T20:40:00Z)
• Dynamic Subgoal-based Exploration via Bayesian Optimization [7.297146495243708]
  Reinforcement learning in sparse-reward navigation environments is challenging and poses a need for effective exploration. We propose a cost-aware Bayesian optimization approach that efficiently searches over a class of dynamic subgoal-based exploration strategies. An experimental evaluation demonstrates that the new approach outperforms existing baselines across a number of problem domains.
  arXiv  Detail & Related papers  (2019-10-21T04:24:29Z)

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.