AutoEnvScaling: Automating the Data Flywheel with Terminal Agents
cs.SE, cs.CL, cs.LG
Submitted: 2026-10-09
Updated: 2026-10-09
Code: https://github.com/harbor-framework/terminal-bench-science
Sources
- Gym-Anything: Turn any Software into an Agent Environment
- SWE-Universe: Scale Real-World Verifiable Environments to Millions
- Terminal-World: Scaling Terminal-Agent Environments via Agent Skills
- Environment Evolution for Terminal Agents
- Endless Terminals: Scaling RL Environments for Terminal Agents
- R-Zero: Self-Evolving Reasoning LLM from Zero Data
- Tmax: A simple recipe for terminal agents
- SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
- Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
- Continual Harness: Online Adaptation for Self-Improving Foundation Agents
- Autodata: An agentic data scientist to create high quality synthetic data
- Meta-Harness: End-to-End Optimization of Model Harnesses
- The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
- Recursive Synthesis for Long-Horizon Terminal Tasks
- SPIRAL: Self-Play on Zero-Sum Games Incentivizes Reasoning via Multi-Agent Multi-Turn Reinforcement Learning
- SPICE: Self-Play In Corpus Environments Improves Reasoning
- SPADE: Self-Play in Adaptive Synthetic Executable Environments
- MLS-Bench: A Holistic and Rigorous Assessment of AI Systems on Building Better AI
- CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks
Related papers
- Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
- GitSkills: A Dataset of Agent Skills on GitHub
- SABER: Benchmarking Operational Safety of LLM Coding Agents in Stateful Project Workspaces
- PackMonitor: Enabling Zero Package Hallucinations Through Decoding-Time Monitoring
- IntentCoding: Amplifying User Intent in Code Generation
- Incentives and Outcomes in Bug Bounties