{"task": {"agent_timeout": 3600, "task": "49", "verifier_timeout": 600, "instruction": "You are given a deep information synthesis question that requires gathering\ndata from multiple web sources and producing a structured JSON answer.\n\nTo analyze progress on Copenhagen Accord commitments, identify the top five EU member states (from those party to the 2009 accord) that achieved the greatest reduction in per capita CO\u2082 emissions.\nThis analysis must be based on the EDGAR (Emissions Database for Global Atmospheric Research) dataset, specifically using figures that exclude LULUCF (Land Use, Land-Use Change, and Forestry).\nCalculate the percentage change in emissions from the 1990 reference year to 2020.\nThe output must be a JSON object, where keys are the country names and values are the calculated percentage change, rounded to one decimal place (e.g., {\"Country\": -15.5}).\n\nResearch this question thoroughly by browsing the web. Find relevant data\nfrom official sources (government databases, statistical offices,\ninternational organizations). Synthesize the information into a single\nJSON answer.\n\nWrite your final answer as a valid JSON dictionary to `/app/answer.json`.\nThe answer should be a JSON object matching the format specified in the\nquestion above (typically string keys with numeric values).\n\nExample answer format:\n```json\n{\"Country A\": 1.23, \"Country B\": 4.56}\n```\n\n**Important:**\n- You should ONLY interact with the environment provided to you AND NEVER ASK FOR HUMAN HELP.\n- Show your work and reasoning before writing the final answer.\n- `/app/answer.json` should contain ONLY the valid JSON dictionary \u2014 no explanation, no markdown fencing.\n", "memory": "", "runnable": false, "difficulty": "difficult", "language": "", "cpus": "", "instruction_truncated": false, "category": "information-synthesis", "compose": false, "has_solution": true, "oracle": null, "docker_image": "", "taskset": "deepsynth", "tags": ["deepsynth", "information-synthesis"]}, "runs": []}