Learning Generalizable Behaviors for Terminal Agents
cs.LG
Submitted: 2026-08-23
Updated: 2026-09-24
Code: https://github.com/laude-institute/harbor
Project page: https://terminal-river.github.io
Terminology
Sources
- Terminal Wrench: A Dataset of 331 Reward-Hackable Environments and 3,632 Exploit Trajectories
- Terminal-World: Scaling Terminal-Agent Environments via Agent Skills
- SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training
- TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks
- Toward Scalable Terminal Task Synthesis via Skill Graphs
- Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRs
- Endless Terminals: Scaling RL Environments for Terminal Agents
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- FrontierSmith: Synthesizing Open-Ended Coding Problems at Scale
- HardTests: Synthesizing High-Quality Test Cases for LLM Coding
- Tmax: A simple recipe for terminal agents
- From Reasoning Traces to Reusable Modules: Understanding Compositional Generalization in Language Model Reasoning
- Recursive Synthesis for Long-Horizon Terminal Tasks
- CLI-Gym: Scalable CLI Task Generation via Agentic Environment Inversion
- On Data Engineering for Scaling LLM Terminal Capabilities
- Rethinking the Trust Region in LLM Reinforcement Learning
- LoCoBench-Agent: An Interactive Benchmark for LLM Agents in Long-Context Software Engineering
- Learning to Reason with Curriculum II: Compositional Generalization
- OpenThoughts-Agent: Data Recipes for Agentic Models
- EvolveCoder: Evolving Test Cases via Adversarial Verification for Code Reinforcement Learning
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks