Convergent Emergence of In-Context Learning Across Modalities

Hugging Face Daily Papers Papers
in-context-learning cross-modality few-shot-learning emergence multi-modal machine-learning

Summary

The paper proposes the Convergent Emergence Hypothesis, stating that few-shot in-context learning emerges with a common cross-modality difficulty profile, and provides empirical support through experiments on six modalities, showing correlated effects in five.

Few-shot in-context learning (ICL), the capacity of a model to infer abstract patterns from input-output examples provided in its prompt and apply them to new inputs, has been extensively studied in large language models trained for next-token prediction on human text. Recently, few-shot ICL has been demonstrated in autoregressive genomic models as well. This raises a question: does ICL emerge broadly across domains, and if so, what common structure is shared? To address both, we develop a controlled cross-modality framework that instantiates the same task suite in a variety of modalities to test what we call the Convergent Emergence Hypothesis: the idea that few-shot ICL, when it emerges, shares a common cross-modality difficulty profile - i.e., tasks that benefit from ICL in one modality tend to benefit in others. We show that paired-mapping ICL emerges across six modalities (language, genome, integer sequences, time series, images, and proteins), surpasses controlled baselines, and has correlated per-task effects across five of them. Together, these results provide support for the Convergent Emergence Hypothesis in some modalities, but not all.
Original Article
View Cached Full Text

Cached at: 09/16/26, 10:46 AM

Paper page - Convergent Emergence of In-Context Learning Across Modalities

Source: https://huggingface.co/papers/2609.14011

Abstract

Few-shotin-contextlearning(ICL),thecapacityofamodeltoinferabstractpatternsfrominput-outputexamplesprovidedinitspromptandapplythemtonewinputs,hasbeenextensivelystudiedinlargelanguagemodelstrainedfornext-tokenpredictiononhumantext.Recently,few-shotICLhasbeendemonstratedinautoregressivegenomicmodelsaswell.Thisraisesaquestion:doesICLemergebroadlyacrossdomains,andifso,whatcommonstructureisshared?Toaddressboth,wedevelopacontrolledcross-modalityframeworkthatinstantiatesthesametasksuiteinavarietyofmodalitiestotestwhatwecalltheConvergentEmergenceHypothesis:theideathatfew-shotICL,whenitemerges,sharesacommoncross-modalitydifficultyprofile-i.e.,tasksthatbenefitfromICLinonemodalitytendtobenefitinothers.Weshowthatpaired-mappingICLemergesacrosssixmodalities(language,genome,integersequences,timeseries,images,andproteins),surpassescontrolledbaselines,andhascorrelatedper-taskeffectsacrossfiveofthem.Together,theseresultsprovidesupportfortheConvergentEmergenceHypothesisinsomemodalities,butnotall.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2609\.14011

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.14011 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.14011 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.14011 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

How Modalities Learn Together (49 minute read)

TLDR AI

A systematic study from Meta FAIR, Reality Labs, and Oxford on multimodal pretraining, revealing asymmetric knowledge flow between modalities, synergy vs. competition dynamics, the benefits of early unification, and efficient training recipes validated with 13.5B MoE models.

Many-Shot CoT-ICL: Making In-Context Learning Truly Learn

Hugging Face Daily Papers

This paper investigates many-shot chain-of-thought in-context learning for reasoning tasks, revealing that standard scaling rules do not transfer and proposing Curvilinear Demonstration Selection (CDS) for improved ordering, achieving up to 5.42 percentage-point gain.

Investigating Cross-Modal Skill Injection: Scenarios, Methods, and Hyperparameters

arXiv cs.CL

This paper systematically investigates cross-modal skill injection, where a domain-expert LLM is merged into a VLM to induce emergent multimodal capabilities. It evaluates different scenarios (instruction-following, cross-lingual, mathematical reasoning), merging methods (TA, DARE, etc.), and hyperparameters, finding that TA and DARE perform well except in mathematical reasoning.

Self-Improving In-Context Learning

arXiv cs.CL

This paper proposes a method to improve in-context learning by optimizing the continuous embeddings of a fixed few-shot prompt at test time, using a self-supervised confidence proxy derived from the model's log-probabilities without requiring fine-tuning or token generation.