✳flâneur — a map of the web's best reading
EWoK: Elements of World Knowledge
ewok-core.github.io · 1,296 words · saved by 1 readers
Elements of World Knowledge (EWoK): A cognition-inspired framework for evaluating basic world knowledge in language models
EWoK: Elements of World Knowledge 🌏 Elements of World Knowledge (EWoK): A cognition-inspired framework for evaluating basic world knowledge in language models --> Anna A Ivanova * 1 , Aalok Sathe * 2 , Benjamin Lipkin * 3 , --> Unnathi Kumar 1 , Setayesh Radkani 3 , Thomas H Clark 3 , Carina Kauf 3 , --> Jennifer Hu 4 , Pramod RT 3 , Gabriel Grand 3 , Vivian Paulun 3 , --> Maria Ryskina 5 , Ekin Akyürek 3 , Ethan Wilcox 6 , Nafisa Rashid 7 , Leshem Choshen 3,8 , --> Roger Levy 3 , Evelina Fedorenko 3 , Joshua Tenenbaum 3 , Jacob Andreas 3 --> 1 Georgia Tech, 2 Brown University, 3 MIT, 4 Johns
Explore this link on the map →related reading
- Language Models, World Models, and Human Model-Buildinglingo.csail.mit.edu
- Explore | alphaXivalphaxiv.org
- LLMs and World Models, Part 1 - by Melanie Mitchellaiguide.substack.com
- Large Language Model: world models or surface statistics?thegradient.pub
- Patterns for Building LLM-based Systems & Productseugeneyan.com
- True Agents Model the Worldprimeintellect.ai
- World Models: Computing the Uncomputablenotboring.co
- Group | Sherry Tongshuang Wucs.cmu.edu
- GitHub - open-compass/opencompass: OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets. · GitHubgithub.com
- [2504.15785] WALL-E 2.0: World Alignment by NeuroSymbolic Learning improves World Model-based LLM Agentsarxiv.org
- LLM Evaluation doesn't need to be complicatedphilschmid.de
- Language Models can Solve Computer Tasksarxiv.org