MARS: Benchmarking the Metaphysical Reasoning Abilities of Language Models with a Multi-task Evaluation Dataset.
Weiqi WangYangqiu SongPublished in: CoRR (2024)
Keyphrases
- language model
- multi task
- language modeling
- multi task learning
- document retrieval
- n gram
- language modelling
- learning tasks
- query expansion
- information retrieval
- probabilistic model
- multiple tasks
- retrieval model
- sparse learning
- multi class
- gaussian processes
- feature selection
- smoothing methods
- statistical language models
- learning problems
- transfer learning
- test collection
- document ranking
- relevance model
- pairwise
- reinforcement learning
- learning algorithm