Ondřej Bojar

Machine translation benchmarks have functionally saturated, but pairing human-authored failure cases with deterministic verification rules reveals critical multimodal blind spots that automated metrics consistently miss.

Pre-pretraining LLMs can yield inconsistent token efficiency gains, revealing the critical influence of experimental setup and random seed on results.

CzechDocs reveals critical insights into format-preserving machine translation, setting the stage for improved translation quality in minority languages.

Generating synthetic training data from limited confidential datasets can produce datasets that are superficially similar to the reference data and improve model training for short answer grading.