# Can In-context Learners Learn a Reasoning Concept from Demonstrations?

arXiv: https://arxiv.org/abs/2212.01692 · HTML: https://mlsanity.com/paper/2212.01692/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 20 of its 27 references that are in the map.
- First posted on arXiv 3 Dec 2022.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 20 of 20)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Scaling Instruction-Finetuned Language Models](https://mlsanity.com/paper/2210.11416/) (2022)
- [Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering](https://mlsanity.com/paper/1809.02789/) (2018)
- [Multitask Prompted Training Enables Zero-Shot Task Generalization](https://mlsanity.com/paper/2110.08207/) (2021)
- [Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?](https://mlsanity.com/paper/2202.12837/) (2022)
- [What Makes Good In-Context Examples for GPT-3?](https://mlsanity.com/paper/2101.06804/) (2021)
- [Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity](https://mlsanity.com/paper/2104.08786/) (2021)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (2022)
- [An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks](https://mlsanity.com/paper/1312.6211/) (2013)
- [Cross-Task Generalization via Natural Language Crowdsourcing Instructions](https://mlsanity.com/paper/2104.08773/) (2021)
- [MetaICL: Learning to Learn In Context](https://mlsanity.com/paper/2110.15943/) (2021)
- [PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts](https://mlsanity.com/paper/2202.01279/) (2022)
- [Improving Question Answering Model Robustness with Synthetic Adversarial Data Generation](https://mlsanity.com/paper/2104.08678/) (2021)
- [WorldTree: A Corpus of Explanation Graphs for Elementary Science Questions supporting Multi-Hop Inference](https://mlsanity.com/paper/1802.03052/) (2018)
- [Towards Ecologically Valid Research on Language User Interfaces](https://mlsanity.com/paper/2007.14435/) (2020)
- [R4C: A Benchmark for Evaluating RC Systems to Get the Right Answer for the Right Reason](https://mlsanity.com/paper/1910.04601/) (2019)

## Built on by (most cited first, 7 of 7)

- [Are Emergent Abilities in Large Language Models just In-Context Learning?](https://mlsanity.com/paper/2309.01809/) (2023)
- [How are Prompts Different in Terms of Sensitivity?](https://mlsanity.com/paper/2311.07230/) (2023)
- [Mimicking or Reasoning: Rethinking Multi-Modal In-Context Learning in Vision-Language Models](https://mlsanity.com/paper/2506.07936/) (2025)
- [Learning 4D Panoptic Scene Graph Generation from Rich 2D Visual Scene](https://mlsanity.com/paper/2503.15019/) (2025)
- [Concept-aware Data Construction Improves In-context Learning of Language Models](https://mlsanity.com/paper/2403.09703/) (2024)
- [Surprise Calibration for Better In-Context Learning](https://mlsanity.com/paper/2506.12796/) (2025)
- [Towards Contamination Resistant Benchmarks](https://mlsanity.com/paper/2505.08389/) (2025)

## Abstract (arXiv)

Language models exhibit an emergent ability to learn a new task from a small number of input-output demonstrations. However, recent work shows that in-context learners largely rely on their pre-trained knowledge, such as the sentiment of the labels, instead of learning new associations from the input. We argue that the commonly-used few-shot evaluation using a random selection of in-context demonstrations can not disentangle models' reliance on such biases, as most of the randomly-selected demonstrations do not present relations informative for prediction beyond exposing the task's input-output distribution. Therefore, to evaluate models' in-context learning ability independent of models' memory, we introduce a Concept-sharing few-shot learning method choosing the demonstrations that share an underlying concept with the predicted sample. We extract a set of such concepts from available human explanations and measure how much models can benefit from presenting these concepts in few-shot demonstrations. We find that most of the recent in-context learners can not consistently benefit from the demonstrated concepts, irrespective of the model size. However, we note that T0 models are more sensitive to exhibited concepts, benefiting from concept-sharing demonstrations in 7 out of 8 evaluation scenarios.