Pith. sign in

REVIEW 5 cited by

Data Distributional Properties Drive Emergent In-Context Learning in Transformers

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2205.05055 v6 pith:MKZL6UZV submitted 2022-04-22 cs.LG cs.AIcs.CL

classification cs.LGcs.AIcs.CL
keywords learningdatain-contexttraininglanguagemodelspropertiesdistributions
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Large transformer-based models are able to perform in-context few-shot learning, without being explicitly trained for it. This observation raises the question: what aspects of the training regime lead to this emergent behavior? Here, we show that this behavior is driven by the distributions of the training data itself. In-context learning emerges when the training data exhibits particular distributional properties such as burstiness (items appear in clusters rather than being uniformly distributed over time) and having large numbers of rarely occurring classes. In-context learning also emerges more strongly when item meanings or interpretations are dynamic rather than fixed. These properties are exemplified by natural language, but are also inherent to naturalistic data in a wide range of other domains. They also depart significantly from the uniform, i.i.d. training distributions typically used for standard supervised learning. In our initial experiments, we found that in-context learning traded off against more conventional weight-based learning, and models were unable to achieve both simultaneously. However, our later experiments uncovered that the two modes of learning could co-exist in a single model when it was trained on data following a skewed Zipfian distribution -- another common property of naturalistic data, including language. In further experiments, we found that naturalistic data distributions were only able to elicit in-context learning in transformers, and not in recurrent models. In sum, our findings indicate how the transformer architecture works together with particular properties of the training data to drive the intriguing emergent in-context learning behaviour of large language models, and how future work might encourage both in-context and in-weights learning in domains beyond language.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. OpenAlex reports about 52 citations worldwide. Full citation record

  1. Can Interpretation Predict Behavior on Unseen Data?

    cs.LG 2025-07 conditional novelty 6.0 of 10

    Presence of hierarchical attention heads on in-distribution data predicts hierarchical out-of-distribution generalization across 270 small transformers, independent of causal support.

  2. Beyond Induction Heads: In-Context Meta Learning Induces Multi-Phase Circuit Emergence

    cs.CL 2025-05 conditional novelty 6.0 of 10

    A two-layer transformer solving an in-context meta-learning task acquires skill in three abrupt phases, each corresponding to a distinct attention circuit: bigram, label attention, then chunking plus label attention.

  3. InSQuAD: In-Context Learning for Efficient Retrieval via Submodular Mutual Information to Enforce Quality and Diversity

    cs.LG 2025-08 conditional novelty 5.0 of 10

    A submodular mutual information framework for selecting and training in-context learning exemplars improves average accuracy on nine benchmarks by about five points over the IDEAL baseline.

  4. Where to show Demos in Your Prompt: A Positional Bias of In-Context Learning

    cs.CL 2025-07 reject novelty 5.0 of 10

    Moving an identical block of in-context examples between prompt sections changes LLM accuracy and prediction stability, with early placements generally outperforming late ones.

  5. Thinking About Thinking: SAGE-nano's Inverse Reasoning for Self-Aware Language Models

    cs.AI 2025-06 reject novelty 3.0 of 10

    A 4B-parameter model is claimed to explain its own reasoning through inverse attention analysis, but the paper offers no consistent evidence or artifacts.

Pith tools