AgentNativeResearchLab/lhtb-glm5.2-trajectories

LHTB trajectories — glm5.2 Detailed agent trajectories on Long-Horizon Terminal-Bench (46 long-horizon terminal tasks, hidden-verifier dense rewards). Harness: LHTB's continue-until-timeout patched Harbor, 90-min budget per task (override_timeout_sec: 5400), n_attempts: 1. Agent: installed CLI agent (glm-52), NOT the paper's Terminus-2 — same tasks/budget/harness, different agent scaffold. Per trial: agent/trajectory.json (ATIF v1.5: steps, tool calls, reasoning, per-step… See the full description on the dataset page: https://huggingface.co/datasets/AgentNativeResearchLab/lhtb-glm5.2-trajector

<iframe src="https://silytics.si/embed/item/6b884ce7-e9d0-4b18-9839-b7df06a918e9" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.