Fix toolorchestrator trajectory loading (#120)

* upload

* update

* Update submission.json

* toolorchestra submission + custom verification logic

* address comments

* Fix toolOrchestrator trajectory loading

- Rename model from ToolOrchestra to Nemotron-Orchestrator-8B
- Update contact info to sdiao@nvidia.com (Shizhe Diao)
- Update trajectory pattern mapping to use new model name

---------

Co-authored-by: hongjin-su <114016954+hongjin-su@users.noreply.github.com>

Author: benshi34

web/leaderboard/src/components/TrajectoryVisualizer.jsx

@@ -141,7 +141,7 @@ const TrajectoryVisualizer = () => {
         'Qwen3-Max-Thinking-Preview': [
           '{domain}_llm_agent_qwen3-max-2025-10-30_user_simulator_gpt-4.1-2025-04-14.json'
         ],
-        'ToolOrchestra': [
+        'Nemotron-Orchestrator-8B': [
           'toolorchestra_{domain}_gpt-5_1trial.json'
         ]
       }