Recursa: Toward Recursively Self-Improving Foundation Model Systems
cs.AI
Submitted: 2026-11-02
Updated: 2026-11-02
License: http://creativecommons.org/licenses/by/4.0/
The gist: Foundation models in changing environments can encounter unfamiliar domain knowledge that their original training did not cover.
Terminology
Abstract
Foundation models in changing environments can encounter unfamiliar domain knowledge that their original training did not cover. However, new domains rarely provide large amounts of expert-created training data or reliable supervisory signals for further training. We introduce Recursa, a foundation model system that adapts to new domains by learning from its own experience without human experts or a stronger teacher model. Recursa uses a knowledge repository to store what it learns: domain-specific skills and instructions on when to apply them, along with memories of past task experiences and reflections on what to improve. Recursa bootstraps the repository from domain materials and then refines it through recursive self-improvement. It uses parallel learners to execute tasks and synthesize local improvements, then validates and merges the improvements into the knowledge repository. On Harvey LAB's Firm Knowledge benchmark, Recursa raises the all-pass rate on held-out test tasks from 30.0% to 55.0% with DeepSeek-V4.1-Flash and from 35.0% to 46.0% with Qwen3.8-Flash.
Sources
- PREPING: Zero-Task Agent Memory Construction via Self-Practice
- SIA: Self Improving AI with Harness & Weight Updates
- Meta-Harness: End-to-End Optimization of Model Harnesses
- Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
- Corpus2Skill: Distilling Enterprise Knowledge into Navigable Agent Skills for QA and RAG
- Harness 2: Recursive Agent Harnessing for an Open World
- ReAct: Synergizing Reasoning and Acting in Language Models
- Hyperagents
- Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models
- RSIAgent: Autonomous Exploration for Recursive Self-improvement in New Environments
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents