Flex-RSI: Benchmarking Recursive Self-Improvement of Frontier Agents
cs.AI, cs.CV, cs.RO
Submitted: 2026-11-06
Updated: 2026-11-06
Code: https://github.com/aiming-lab/RSI-Exam
Project page: https://flex-rsi.github.io
Terminology
Sources
- The Surprising Effectiveness of Test-Time Training for Few-Shot Learning
- On Evaluation of Embodied Navigation Agents
- ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence
- Qwen3-VL Technical Report
- Qwen2.5-VL Technical Report
- RoboHarn-Evo: Evolving Hierarchical Physical Knowledge for Self-Improving Robotic Manipulation
- ARKitScenes: A Diverse Real-World Dataset For 3D Indoor Scene Understanding Using Mobile RGB-D Data
- Audit the Scaffold, Not the Checkpoint: A Stationarity Dichotomy for Recursive Self-Improvement in Agentic Coding
- Benchmark Designers Should "Train on the Test Set" to Expose Exploitable Non-Visual Shortcuts
- Trial by Trace: Benchmarking AI Agent Harnesses for Reliability and Failure Attribution
- Recursive Self-Improvement in AI: From Bounded Self-Refinement to Autonomous Research Loops
- VoLo: A Physical Orchestrator for Open-Vocabulary Long-Horizon Manipulation
- RoboDojo: A Unified Sim-and-Real Benchmark for Comprehensive Evaluation of Generalist Robot Manipulation Policies
- Are Self-Evolving Agents Continual Learners?
- Learning to Compose: Recursive Self-Improvement of Shared Executable Stacks for Zero-Shot Long-Horizon Robot Manipulation
- Auditing the Product, Not the Recursion: The Evaluation Evidence Behind Recursive Self-Improvement Claims
- On the Measure of Intelligence
- ARC Prize 2024: Technical Report
- ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems
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