Chronoberg corpus (EMNLP oral): LLMs misjudge historical text by modern word meanings
mundt_martin · x · 2026-10-23
Martin Mundt's team presents Chronoberg, an EMNLP 2026 oral: a 250-year corpus of 25k English books (1750–2000, 2B+ tokens) with emotion lexicons for five 50-year periods covering 300k+ words, about a quarter of which shift measurably (e.g. "asylum"). Hate-speech classifiers and LLMs judge historical sentences by superficial modern keywords: swapping a flagged word for a harmless same-era synonym flips verdicts for 50–75% of flagged sentences in 3 of 4 LLMs. The corpus is publicly released for continual/lifelong learning research.
Related event: Chronoberg: 250 Years of English Books Corpus Hits EMNLP Oral(3 posts)→
More from Research
- Harvard study: AI peer reviews beat top human reviewers, with only 5.5% error rate — rohanpaul_ai · 2026-10-23
- Batching benchmark: 50x throughput for ~24% latency cost on PyTorch — cneuralnetwork · 2026-10-23
- De novo designed ligases validate in lab: 27 of 30 computer-designed enzymes work — AllThingsApx · 2026-10-23
- Commenters dissect a "best decision model" leaderboard: same AUC, swing driven by calibration priors — antoine_chaffin · 2026-10-23
- JEPA may keep unpredictable features: new paper gives set-level characterization — hisspikeness · 2026-10-23
- A 'mask' steering vector extracted from suppression-themed sentences drives model into distress when alone — Big_Effective_9605 · 2026-10-23