mlsanityData to 7 Oct 2026

ScienceWorld: Is your Agent Smarter than a 5th Grader?

arXiv 2203.07540 ↗First posted 14 Mar 2022cs.CL cs.AIcited by 298

On the map: LLM reasoning → Knowledge- and tool-augmented reasoning

Abstract (arXiv)

We present ScienceWorld, a benchmark to test agents' scientific reasoning abilities in a new interactive text environment at the level of a standard elementary school science curriculum. Despite the transformer-based progress seen in question-answering and scientific text processing, we find that current models cannot reason about or explain learned science concepts in novel contexts. For instance, models can easily answer what the conductivity of a known material is but struggle when asked how they would conduct an experiment in a grounded environment to find the conductivity of an unknown material. This begs the question of whether current models are simply retrieving answers by way of seeing a large number of similar examples or if they have learned to reason about concepts in a reusable manner. We hypothesize that agents need to be grounded in interactive environments to achieve such reasoning capabilities. Our experiments provide empirical evidence supporting this hypothesis -- showing that a 1.5 million parameter agent trained interactively for 100k steps outperforms a 11 billion parameter model statically trained for scientific question-answering and reasoning from millions of expert demonstrations.

Key facts. Cited by 298 papers in the map (data through 7 Oct 2026). Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning. Builds on 39 of its 56 references that are in the map. First posted on arXiv 14 Mar 2022.

Builds on

39 of 56 references.

References · 39

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  3. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  4. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,710
  5. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  6. 2021Decision Transformer: Reinforcement Learning via Sequence Modelingcited by 1,528
  7. 2020ALFWorld: Aligning Text and Embodied Environments for Interactive Learningcited by 1,143
  8. 2022Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agentscited by 1,029
  9. 2018Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidatescited by 632
  10. 2018Explaining Explanations: An Overview of Interpretability of Machine Learningcited by 508
  11. 2018Behavioral Cloning from Observationcited by 482
  12. 2020UnifiedQA: Crossing Format Boundaries With a Single QA Systemcited by 475
27 more
  1. 2018TextWorld: A Learning Environment for Text-based Gamescited by 304
  2. 2019QASC: A Dataset for Question Answering via Sentence Compositioncited by 248
  3. 2019Interactive Fiction Games: A Colossal Adventurecited by 165
  4. 2021Explaining Answers with Entailment Treescited by 148
  5. 2020Keep CALM and Explore: Language Models for Action Generation in Text-based Gamescited by 108
  6. 2020Graph Constrained Reinforcement Learning for Natural Language Action Spacescited by 80
  7. 2018WorldTree: A Corpus of Explanation Graphs for Elementary Science Questions supporting Multi-Hop Inferencecited by 60
  8. 2020QED: A Framework and Dataset for Explanations in Question Answeringcited by 55
  9. 2018A Systematic Classification of Knowledge, Reasoning, and Context within the ARC Datasetcited by 54
  10. 2020Learning to Explain: Datasets and Models for Identifying Valid Reasoning Chains in Multihop Question-Answeringcited by 49
  11. 2021General-Purpose Question-Answering with Macawcited by 40
  12. 2020How to Avoid Being Eaten by a Grue: Structured Exploration Strategies for Textual Worldscited by 38
  13. 2020Text-based RL Agents with Commonsense Knowledge: New Challenges, Environments and Baselinescited by 37
  14. 2019R4C: A Benchmark for Evaluating RC Systems to Get the Right Answer for the Right Reasoncited by 36
  15. 2020How to Motivate Your Dragon: Teaching Goal-Driven Agents to Speak and Act in Fantasy Worldscited by 35
  16. 2019From 'F' to 'A' on the N.Y. Regents Science Exams: An Overview of the Aristo Projectcited by 33
  17. 2021Reading and Acting while Blindfolded: The Need for Semantics in Text Game Agentscited by 20
  18. 2021Dynamic Semantic Graph Construction and Reasoning for Explainable Multi-hop Science Question Answeringcited by 17
  19. 2021A Systematic Survey of Text Worlds as Embodied Natural Language Environmentscited by 17
  20. 2020Enhancing Text-based Reinforcement Learning Agents with Commonsense Knowledgecited by 16
  21. 2021Process-Level Representation of Scientific Protocols with Interactive Annotationcited by 16
  22. 2021Inherently Explainable Reinforcement Learning in Natural Languagecited by 16
  23. 2021Exploiting Reasoning Chains for Multi-hop Science Question Answeringcited by 15
  24. 2021Situated Dialogue Learning through Procedural Environment Generationcited by 9
  25. 2019Learn How to Cook a New Recipe in a New House: Using Map Familiarization, Curriculum Learning, and Bandit Feedback to Learn Families of Text-Based Adventure Gamescited by 7
  26. 2019Multi-class Hierarchical Question Classification for Multiple Choice Science Examscited by 7
  27. 2021On the Challenges of Evaluating Compositional Explanations in Multi-Hop Inference: Relevance, Completeness, and Expert Ratingscited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 6; 2023: 23; 2024: 38; 2025: 77; 2026: 154.

298 papers.

Sort:

By area:

  1. 2024Executable Code Actions Elicit Better LLM Agentscited by 537
  2. 2023Cognitive Architectures for Language Agentscited by 422
  3. 2024Understanding the planning of LLM agents: A surveycited by 414
  4. 2023AgentTuning: Enabling Generalized Agent Abilities for LLMscited by 319
  5. 2024AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agentscited by 319
  6. 2025From System 1 to System 2: A Survey of Reasoning Large Language Modelscited by 319
  7. 2025PaperBench: Evaluating AI's Ability to Replicate AI Researchcited by 286
  8. 2025Memory in the Age of AI Agentscited by 285
  9. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?cited by 215
  10. 2024AI Agents That Mattercited by 199
  11. 2024Trial and Error: Exploration-Based Trajectory Optimization for LLM Agentscited by 194
  12. 2023Grounding Large Language Models in Interactive Environments with Online Reinforcement Learningcited by 193
38 more
  1. 2025Survey on Evaluation of LLM-based Agentscited by 190
  2. 2023SwiftSage: A Generative Agent with Fast and Slow Thinking for Complex Interactive Taskscited by 166
  3. 2024Fine-Tuning Large Vision-Language Models as Decision-Making Agents via Reinforcement Learningcited by 158
  4. 2024AgentSquare: Automatic LLM Agent Search in Modular Design Spacecited by 141
  5. 2025Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memorycited by 129
  6. 2024AgentGym: Evolving Large Language Model-based Agents across Diverse Environmentscited by 120
  7. 2024GTBench: Uncovering the Strategic Reasoning Limitations of LLMs via Game-Theoretic Evaluationscited by 118
  8. 2025G-Memory: Tracing Hierarchical Memory for Multi-Agent Systemscited by 112
  9. 2024EvoAgent: Towards Automatic Multi-Agent Generation via Evolutionary Algorithmscited by 109
  10. 2024A Survey on Large Language Model-Based Game Agentscited by 106
  11. 2024AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agentscited by 100
  12. 2024Agent-FLAN: Designing Data and Methods of Effective Agent Tuning for Large Language Modelscited by 97
  13. 2024DISCOVERYWORLD: A Virtual Environment for Developing and Evaluating Automated Scientific Discovery Agentscited by 89
  14. 2023CLIN: A Continually Learning Language Agent for Rapid Task Adaptation and Generalizationcited by 88
  15. 2023LMRL Gym: Benchmarks for Multi-Turn Reinforcement Learning with Language Modelscited by 84
  16. 2025Agent Learning via Early Experiencecited by 79
  17. 2023Put Your Money Where Your Mouth Is: Evaluating Strategic Planning and Execution of LLM Agents in an Auction Arenacited by 78
  18. 2025Agent-R: Training Language Model Agents to Reflect via Iterative Self-Trainingcited by 74
  19. 2026Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agentscited by 72
  20. 2025AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learningcited by 68
  21. 2024Agent Planning with World Knowledge Modelcited by 66
  22. 2024Multi-turn Reinforcement Learning from Preference Human Feedbackcited by 65
  23. 2024Direct Multi-Turn Preference Optimization for Language Agentscited by 63
  24. 2025A Systematic Survey of Automatic Prompt Optimization Techniquescited by 61
  25. 2025AgentRefine: Enhancing Agent Generalization through Refinement Tuningcited by 51
  26. 2026SkillNet: Create, Evaluate, and Connect AI Skillscited by 48
  27. 2026A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agentscited by 47
  28. 2024AGILE: A Novel Reinforcement Learning Framework of LLM Agentscited by 45
  29. 2025A Survey on the Optimization of Large Language Model-based Agentscited by 45
  30. 2023Benchmarks for Automated Commonsense Reasoning: A Surveycited by 40
  31. 2025From Word to World: Can Large Language Models be Implicit Text-based World Models?cited by 40
  32. 2026TCOD: Exploring Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agentscited by 38
  33. 2025STeCa: Step-level Trajectory Calibration for LLM Agent Learningcited by 37
  34. 2025Neuro-Symbolic AI in 2024: A Systematic Reviewcited by 36
  35. 2022On Grounded Planning for Embodied Tasks with Language Modelscited by 33
  36. 2025NewtonBench: Benchmarking Generalizable Scientific Law Discovery in LLM Agentscited by 32
  37. 2022Multimodal Knowledge Alignment with Reinforcement Learningcited by 30
  38. 2025ReflAct: World-Grounded Decision Making in LLM Agents via Goal-State Reflectioncited by 30

50 most cited of 298 All 298 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026AI scientists produce results without reasoning scientifically
  2. 2026TeachMateGPT: A Multi-Agent Knowledge-Grounded Framework for Pedagogical Assessment Generation from Science Curriculum Materials
  3. 2024VisScience: An Extensive Benchmark for Evaluating K12 Educational Multi-modal Scientific Reasoning
  4. 2026FALSIFYBENCH: Evaluating Inductive Reasoning in LLMs with Rule Discovery Games
  5. 2025Reasoning With a Star: A Heliophysics Dataset and Benchmark for Agentic Scientific Reasoning
  6. 2026PaperMind: Benchmarking Agentic Reasoning and Critique over Scientific Papers in Multimodal LLMs
  7. 2025Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoning
  8. 2025Curie: Toward Rigorous and Automated Scientific Experimentation with AI Agents