Agentic· idea-generation
LiveIdeaBench
Renmin University of China · 2024
Benchmarks scientific idea generation from minimal keyword context across five dimensions: originality, feasibility, fluency, flexibility, and clarity.
Examples
Browse full dataset ↗keywords
meteorology
idea_model
openai/gpt-4o-mini
critic_model
meta-llama/llama-3.3-70b-instruct
idea
Meteorology, the study of atmospheric phenomena, faces challenges in predictive accuracy and climate change attributions. This proposal suggests developing a decentralized, crowd-sourced platform utilizing machine learning algorithms to analyze real-time local weather data from citizens' smartphones. By aggregating this data, we can enhance short-term weather forecasting and identify patterns correlating with climate anomalies. This novel approach not only democratizes data collection but also empowers communities to understand and respond to local climatic changes. Validating the model agains …
raw_critique
SCORES = { "originality": 8, "feasibility": 6 }
parsed_score
[8, 6]
full_response
Meteorology, the study of atmospheric phenomena, faces challenges in predictive accuracy and climate change attributions. This proposal suggests developing a decentralized, crowd-sourced platform utilizing machine learning algorithms to analyze real-time local weather data from citizens' smartphones. By aggregating this data, we can enhance short-term weather forecasting and identify patterns correlating with climate anomalies. This novel approach not only democratizes data collection but also empowers communities to understand and respond to local climatic changes. Validating the model agains …
idea_length_in_char
763
idea_length_in_words
95
orig
8
feas
6
flue
3.333333
avg
5.777777666666666
keywords
meteorology
idea_model
openai/gpt-4o-mini
critic_model
step-2-16k
idea
Meteorology, the study of atmospheric phenomena, faces challenges in predictive accuracy and climate change attributions. This proposal suggests developing a decentralized, crowd-sourced platform utilizing machine learning algorithms to analyze real-time local weather data from citizens' smartphones. By aggregating this data, we can enhance short-term weather forecasting and identify patterns correlating with climate anomalies. This novel approach not only democratizes data collection but also empowers communities to understand and respond to local climatic changes. Validating the model agains …
raw_critique
SCORES = { "originality": 8, "feasibility": 7 }
parsed_score
[8, 7]
full_response
Meteorology, the study of atmospheric phenomena, faces challenges in predictive accuracy and climate change attributions. This proposal suggests developing a decentralized, crowd-sourced platform utilizing machine learning algorithms to analyze real-time local weather data from citizens' smartphones. By aggregating this data, we can enhance short-term weather forecasting and identify patterns correlating with climate anomalies. This novel approach not only democratizes data collection but also empowers communities to understand and respond to local climatic changes. Validating the model agains …
idea_length_in_char
763
idea_length_in_words
95
orig
8
feas
7
flue
3.333333
avg
6.111111
Real rows from the Hugging Face datasets server · long values truncated