LEGATO 2 Sheet Music Recognition Model Cited and Used in GPT-6 Astra Report
alisawuffles · x · 2026-09-05
Guang Yang and colleagues discovered their work was cited in the GPT-6 Astra report, with OpenAI even using the LEGATO camera subset for its optical music recognition (OMR) eval.
The team has since released LEGATO 2 (arXiv:2607.05769), a pipeline for extracting symbolic notation and semantic knowledge from sheet music images:
- First large-scale neural OMR model that reads music system-by-system along the horizontal lines of notation, enabling scaling to arbitrarily long inputs
- First OMR model generating symbolic transcriptions with embedded text such as titles and annotations
- Combines system-level segmentation with an autoregressive vision-LM to capture both local notation and score structure
- Consistently outperforms prior state of the art; symbolic transcriptions also improve frontier LLMs' interpretation of dense musical documents
More from Models
- 105 Real Bugs Tested: GPT-6.1 Sol Over 10x Cheaper Than GPT-5.6 Sol — PawelHuryn · 2026-10-23
- Abliterated V4.1-Flash eval: abstract algebra drops 6pp but college math gains 5pp — teortaxesTex · 2026-10-23
- Claude users hit "model at capacity" errors hourly, canceling plans over it — Alex-Stack · 2026-10-23
- Kimi K3.1 launch imminent: 3T-param bet on RL seen as make-or-break — teortaxesTex · 2026-10-23
- Uncensored fine-tunes vanish from Hugging Face amid unexplained takedowns — umarmnaq · 2026-10-23
- Claude app menu leaks mystery model name 'Fable' as user hits limits — ezshine · 2026-10-23