inflection
opengloss-v2.1-inflections
Superseded by OpenGloss v2.2 (2026-09-08): 148,292 live lexemes and 288,304 senses — tier 5 closes the WordNet gap (38,100 entries imported from Princeton WordNet 3.0 and enriched), inflected-form headwords are folded onto their lemmas, and every inherited field carries a migrate provenance record. v2.1 stays published for reproducibility.
OpenGloss v2.1 — Inflections
A flat form→lemma lookup table, one row per surface string a consumer might actually type or scan: every… See the full description on the dataset page: https://huggingface.co/datasets/mjbommar/opengloss-v2.1-inflections.opengloss-v2.2-inflections
Superseded by OpenGloss v2.3 (2026-09-09): tier 6 adds ~12,000 named entities (people, places, organizations, works, events) with entity_type, Wikidata ids and alias_of links, and every proper noun in the release is now typed. v2.2 stays published for reproducibility.
OpenGloss v2.2 — Inflections
A flat form→lemma lookup table, one row per surface string a consumer might actually type or scan: every stored inflected form (plural, past_tense, past_participle, present_participle… See the full description on the dataset page: https://huggingface.co/datasets/mjbommar/opengloss-v2.2-inflections.opengloss-v2.3-inflections
OpenGloss v2.3 — Inflections
A flat form→lemma lookup table, one row per surface string a consumer might actually type or scan: every stored inflected form (plural, past_tense, past_participle, present_participle, third_person_singular, comparative, superlative), every recorded derivation, and — critically — one lemma row for the headword itself, so resolving any surface string, inflected or not, is the same one lookup rather than a branch on whether stemming is needed first.… See the full description on the dataset page: https://huggingface.co/datasets/mjbommar/opengloss-v2.3-inflections.inflection_etInflection-BenchmarksCloned from https://github.com/InflectionAI/Inflection-Benchmarks
MT-Bench Inf
In mt_bench_inf.jsonl we release a corrected version of the MT-Bench questions that we use for evaluation. Each entry has the following fields:
question_id: The question number
category: Which MT-Bench category
turn: A list with the turns
reference [optional]: A reference answer
Below, we show a few examples of questions, the original GPT-4 Reference answer, and our corrected answer:… See the full description on the dataset page: https://huggingface.co/datasets/academic-datasets/Inflection-Benchmarks.icelandic-inflection-medium