Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?
Large language models (LMs) are able to in-context learn -- perform a new task via inference alone by conditioning on a few input-label pairs (demonstrations) and making predictions for new inputs. However, there has been little understanding of how the model learns and which aspects of the demonstrations contribute to end task performance. In this paper, we show that ground truth demonstrations are in fact not required -- randomly replacing labels in the demonstrations barely hurts performance on a range of classification and multi-choce tasks, consistently over 12 different models including GPT-3. Instead, we find that other aspects of the demonstrations are the key drivers of end task performance, including the fact that they provide a few examples of (1) the label space, (2) the distribution of the input text, and (3) the overall format of the sequence. Together, our analysis provides a new way of understanding how and why in-context learning works, while opening up new questions about how much can be learned from large language models through inference alone.
Code (2)
Tasks
In-Context LearningMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
When Correct Demonstrations Hurt: Rethinking the Role of Exemplars in In-Context Learning
In-context learning (ICL) is often motivated by the intuition that demonstrations help because they provide correct input-output examples. However, we reveal a counterintuitive phenomenon: correctness does not guarantee …
Logical ReasoningWhat In-Context Learning "Learns" In-Context: Disentangling Task Recognition and Task Learning
Large language models (LLMs) exploit in-context learning (ICL) to solve tasks with only a few demonstrations, but its mechanisms are not yet well-understood. Some works suggest that LLMs only recall already learned conce…
In-Context LearningRethinking Label Consistency of In-Context Learning: An Implicit Transductive Label Propagation Perspective
Large language models (LLMs) perform in-context learning (ICL) with minimal supervised examples, which benefits various natural language processing (NLP) tasks. One of the critical research focus is the selection of prom…
Rethinking and Recomputing the Value of ML Models
In this paper, we argue that the way we have been training and evaluating ML models has largely forgotten the fact that they are applied in an organization or societal context as they provide value to people. We show tha…
What Do Safety-Aligned LLMs Learn From Mixed Compliance Demonstrations?
Prior work has shown that in-context demonstrations can jailbreak language models, but it remains unclear how models interpret different types of compliance demonstrations. We study this by mixing benign compliance demon…