ICL emerges across language, genomes, proteins, and more — the Convergent Emergence Hypothesis
DanielKhashabi · x · 2026-09-17
A new paper tests whether in-context learning is unique to LLMs by training models on six very different data types — language, genomes, proteins, integer sequences, time series, and images — on the same abstract tasks:
- Few-shot ICL emerges in all six modalities
- Surprisingly, across five modalities the same tasks tend to be easy or hard to learn in context, despite distinct per-modality strengths
- The authors propose the Convergent Emergence Hypothesis: ICL arises from next-token prediction across rich, structured data, with a shared core transcending modality
Work led by Nate Schlueter.
Related event: JHU study finds in-context learning emerges across six modalities(4 posts)→
More from Research
- Radial gate objective cuts diffusion transformer FID from 16.24 to 11.83 — Junha Park · 2026-11-02
- REA: async brain-cerebellum runtime for robots where policies never wait on the LLM — Pengfei Zhou · 2026-11-02
- Microsoft's MemSteer turns unlabeled user histories into agent memory supervision — MicrosoftResearch · 2026-11-02
- SciGame benchmark uses counterfactual worlds to expose LLM agents' priors vs evidence — Yusheng Zhao · 2026-11-02
- On Terence Tao's blog, Frank Calegari calls on mathematicians to adapt after OpenAI's 700-paper release — Hesamation · 2026-11-02
- Tsinghua & Microsoft's HarnessMM cuts agent token use by up to 87.93% while raising success rate — rohanpaul_ai · 2026-11-02