# The Stochastic Parrot on LLM's Shoulder: A Summative Assessment of Physical Concept Understanding

arXiv: https://arxiv.org/abs/2502.08946 · HTML: https://mlsanity.com/paper/2502.08946/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 37 of its 52 references that are in the map.
- First posted on arXiv 13 Feb 2025.

## Builds on (most cited first, 30 of 37)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [Fully Convolutional Networks for Semantic Segmentation](https://mlsanity.com/paper/1411.4038/) (2014)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021)
- [Mistral 7B](https://mlsanity.com/paper/2310.06825/) (2023)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Rich feature hierarchies for accurate object detection and semantic segmentation](https://mlsanity.com/paper/1311.2524/) (2013)
- [A Survey on Evaluation of Large Language Models](https://mlsanity.com/paper/2307.03109/) (2023)
- [Habitat: A Platform for Embodied AI Research](https://mlsanity.com/paper/1904.01201/) (2019)
- [How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites](https://mlsanity.com/paper/2404.16821/) (2024)
- [Recent Advances in Natural Language Processing via Large Pre-Trained Language Models: A Survey](https://mlsanity.com/paper/2111.01243/) (2021)
- [The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"](https://mlsanity.com/paper/2309.12288/) (2023)
- [A Survey of Embodied AI: From Simulators to Research Tasks](https://mlsanity.com/paper/2103.04918/) (2021)
- [Large Language Models as General Pattern Machines](https://mlsanity.com/paper/2307.04721/) (2023)
- [The Debate Over Understanding in AI's Large Language Models](https://mlsanity.com/paper/2210.13966/) (2022)
- [Language Models Meet World Models: Embodied Experiences Enhance Language Models](https://mlsanity.com/paper/2305.10626/) (2023)
- [Hypothesis Search: Inductive Reasoning with Language Models](https://mlsanity.com/paper/2309.05660/) (2023)
- [Do Androids Laugh at Electric Sheep? Humor "Understanding" Benchmarks from The New Yorker Caption Contest](https://mlsanity.com/paper/2209.06293/) (2022)
- [Flocks of Stochastic Parrots: Differentially Private Prompt Learning for Large Language Models](https://mlsanity.com/paper/2305.15594/) (2023)
- [LLMs and the Abstraction and Reasoning Corpus: Successes, Failures, and the Importance of Object-based Representations](https://mlsanity.com/paper/2305.18354/) (2023)
- [FLUTE: Figurative Language Understanding through Textual Explanations](https://mlsanity.com/paper/2205.12404/) (2022)
- [Communicating Natural Programs to Humans and Machines](https://mlsanity.com/paper/2106.07824/) (2021)
- [Metaphor Understanding Challenge Dataset for LLMs](https://mlsanity.com/paper/2403.11810/) (2024)
- [The Dark Side of ChatGPT: Legal and Ethical Challenges from Stochastic Parrots and Hallucination](https://mlsanity.com/paper/2304.14347/) (2023)
- [Personality Understanding of Fictional Characters during Book Reading](https://mlsanity.com/paper/2305.10156/) (2023)

## Built on by (most cited first, 11 of 11)

- [Exploring the Evolution of Physics Cognition in Video Generation: A Survey](https://mlsanity.com/paper/2503.21765/) (2025)
- [PRELUDE: A Benchmark Designed to Require Global Comprehension and Reasoning over Long Contexts](https://mlsanity.com/paper/2508.09848/) (2025)
- [Truly Assessing Fluid Intelligence of Large Language Models through Dynamic Reasoning Evaluation](https://mlsanity.com/paper/2506.02648/) (2025)
- [Surfacing Subtle Stereotypes: A Multilingual, Debate-Oriented Evaluation of Modern LLMs](https://mlsanity.com/paper/2511.01187/) (2025)
- [Many-Shot CoT-ICL: Making In-Context Learning Truly Learn](https://mlsanity.com/paper/2605.13511/) (2026)
- [Ref-Long: Benchmarking the Long-context Referencing Capability of Long-context Language Models](https://mlsanity.com/paper/2507.09506/) (2025)
- [InPhyRe Discovers: Large Multimodal Models Struggle in Inductive Physical Reasoning](https://mlsanity.com/paper/2509.12263/) (2025)
- [Beyond Mimicry: Preference Coherence in LLMs](https://mlsanity.com/paper/2511.13630/) (2025)
- [ClimAgent: LLM as Agents for Autonomous Open-ended Climate Science Analysis](https://mlsanity.com/paper/2604.16922/) (2026)
- [GPT and Prejudice: A Sparse Approach to Understanding Learned Representations in Large Language Models](https://mlsanity.com/paper/2510.01252/) (2025)
- [InsightEmb: Learning Action-Intent Embeddings for Agentic Insight Retrieval](https://mlsanity.com/paper/2608.04761/) (2026)

## Abstract (arXiv)

In a systematic way, we investigate a widely asked question: Do LLMs really understand what they say?, which relates to the more familiar term Stochastic Parrot. To this end, we propose a summative assessment over a carefully designed physical concept understanding task, PhysiCo. Our task alleviates the memorization issue via the usage of grid-format inputs that abstractly describe physical phenomena. The grids represents varying levels of understanding, from the core phenomenon, application examples to analogies to other abstract patterns in the grid world. A comprehensive study on our task demonstrates: (1) state-of-the-art LLMs, including GPT-4o, o1 and Gemini 2.0 flash thinking, lag behind humans by ~40%; (2) the stochastic parrot phenomenon is present in LLMs, as they fail on our grid task but can describe and recognize the same concepts well in natural language; (3) our task challenges the LLMs due to intrinsic difficulties rather than the unfamiliar grid format, as in-context learning and fine-tuning on same formatted data added little to their performance.