🖥 A mere 9B open model rivals Claude Haiku 4.5 on terminal-use benchmarks. And the data, models, and code are all open. Title: Tmax: A simple recipe for terminal agents URL: An Ai2/UW work that gives terminal-agent RL training a simple, reproducible recipe. Here are the 3 standout points. 🧩 A compositional data generation pipeline Tasks are synthesized at scale from the product of 9 axes (domain, skills, verifier, persona, difficulty, and more), with Gemini-3-Pro generating the Dockerfile and unit tests. Expensive quality validation is skipped; RL just drops pass-rate-0 samples. The result is Tmax-15K, 14,600 environments, over 2.5x larger than prior data, contamination-free and the hardest of the bunch. ⚙️ A simple outcome-only RL recipe The reward is just "did it complete the task." Since naive GRPO collapses in long-horizon agent settings, they stabilize with DPPO (masking divergent logprobs), an FP32 LM head, and a large group size of 32. This recipe lifts Qwen 3.5 9B to 27% on Terminal-Bench 2.0. 📈 Strength, generalization, and full release It improves over the Qwen baseline at every size from 2B to 27B. It also lifts non-terminal evals, SWE-Bench Verified 44 to 53.5 and AIME 73 to 91, and generalizes across harnesses and model families. Data, models, and code are fully open on GitHub. A strong foundation for open terminal-agent research. #TerminalAgents# #ReinforcementLearning#

Nice little survey on Terminal Agents. It provides good information on what exactly is a terminal agent, and why do harness comparisons keep contradicting each other? Paper: Track more trending AI papers in our academy:

Training data for terminal agents often ships tasks where instruction, environment, solution, and verifier disagree — producing unsolvable tasks. A new synthesis framework cuts that at the root. Title: FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis URL: 📌 Overview Reconstructs related skills into rich scenarios, builds the environment first, and grounds instruction, solution, and verifier in that same executable state. 🧩 Problems it solves ・Multi-stage generation loses source dependencies and intermediate states ・Misaligned instruction/environment/solution/verifier yield unsolvable, unverifiable tasks ⚙️ Method ・Collects 71K+ skills, reconstructs into 5-dimensional scenarios ・Builds the environment in Docker, exposing its real state as a shared channel ・Generates instruction→solution→verifier sequentially; a router pinpoints and repairs only the failing part 📊 Results ・Synthesizes 6,078 validated tasks at 22.77 tests/task on average ・Fine-tuning Qwen3.5: 4B +40.5%, 9B +30.1%, 27B +16.5% ・The 27B (47.57) nears the ~15x larger 397B (49.06) ・70% end-to-end yield vs 15–28% for competitors A clear case that high-quality executable tasks come from careful state grounding, not brute-force generation. #LLMAgents# #TerminalBench#

Tencent HY proposes environment evolution, which builds off-policy lineages of increasingly difficult verified terminal tasks, keeping long-horizon RL signals alive as Qwen3.6 agents improve. Environment Evolution for Terminal Agents Paper: