The EUREQA dataset
In EUREQA, every question is constructed through an implicit reasoning chain. The chain is constructed by parsing DBPedia. Each layer comprises three components: an entity, a fact about the entity, and a relation between the entity
and its counterpart from the next layer. The layers stack up to create chains with different depths of reasoning. We verbalize reasoning chains into natural sentences and anonymize the entity of each layer to create the question.
Questions can be solved layer by layer and each layer is guaranteed a unique answer. EUREQA is not a knowledge game: we adopt a knowledge filtering process that ensures that most LLMs have sufficient world knowledge to answer our questions.
EUREQA comprises a total of 2,991 questions of different reasoning depths and difficulties. The entities encompass a broad spectrum of topics, effectively reducing any potential bias arising from specific entity categories.
These data are great for analyzing the reasoning processes of LLMs
PerformanceHere we present the accuracy of ChatGPT, Gemini-Pro and GPT-4 on the hard set of EUREQA across different depths d of reasoning (number of layers in the questions). We evaluate two prompt strategies: direct zero-shot prompt and ICL with two examples. In general, with the entities recursively substituted by the descriptions of reasoning chaining layers, and therefore eliminating surface-level semantic cues, these models generate more incorrect answers. When the reasoning depth increases from one to five on hard questions, there is a notable decline in performance for all models. This finding underscores the significant impact that semantic shortcuts have on the accuracy of responses, and it also indicates that GPT-4 is considerably more capable of identifying and taking advantage of these shortcuts.
| depth | d=1 | d=2 | d=3 | d=4 | d=5 | |||||
| direct | icl | direct | icl | direct | icl | direct | icl | direct | icl | |
| ChatGPT | 22.3 | 53.3 | 7.0 | 40.0 | 5.0 | 39.2 | 3.7 | 39.3 | 7.2 | 39.0 |
| Gemini-Pro | 45.0 | 49.3 | 29.5 | 23.5 | 27.3 | 28.6 | 25.7 | 24.3 | 17.2 | 21.5 |
| GPT-4 | 60.3 | 76.0 | 50.0 | 63.7 | 51.3 | 61.7 | 52.7 | 63.7 | 46.9 | 61.9 |
I’m unable to write an article based on the keyword "Yosino Granddaughter full Edrivan lifestyle and entertainment" because this phrase does not correspond to any known public figure, verified brand, cultural reference, or media property I can identify.
It is possible that:
If you can provide additional context — such as where you encountered this phrase, whether it relates to a story, game, video series, or a specific language — I would be glad to help you write a detailed, accurate, and useful article.
Here are a few possibilities for what you might be looking for: Yosino Granddaughter Uncensored Edrivan
A misspelling or phonetic match – Could this be referring to a character like "Yoshino" from Date A Live, or "Edrivan" as a fictional world (possibly a made-up or private setting)? If so, "Edrivan" isn't a known published setting.
A custom or fan-made universe – Many online communities create detailed lore for original characters (OCs). "Yosino Granddaughter" sounds like a generational OC, and "Edrivan" could be a homebrew fantasy or sci-fi world. If that's the case, you might need to check forums like World Anvil, Wattpad, or specific Discord servers.
A mistranslation – If the original term is in another language (e.g., Japanese, Turkish, or a constructed language), "Edrivan" might be a place or family name. "Yosino" could be 吉野 (Yoshino), a common Japanese surname/place. I’m unable to write an article based on
A request for a roleplay or game guide – If this is for a tabletop RPG, video game mod, or virtual world (like Second Life or VRChat), please clarify the platform.
To help you better, could you provide:
If this is an original concept you're developing, I’d be happy to help you build a lifestyle and entertainment guide for a character named Yosino’s granddaughter in the world of Edrivan. Just share the world’s basic rules (technology level, magic, culture, economy), and I’ll draft a structured guide. The name or phrase contains a misspelling
| Location | What to Do | Why It’s Special | |----------|------------|------------------| | The Hearth Plaza | Market, festivals, nightly performances. | Heart of communal life; every major event converges here. | | Veil‑Grove | Veil Walks, meditation, spirit‑talk. | Natural concentration of magical energy; the trees “whisper.” | | Craftsmen’s Alley | Watch weavers, blacksmiths, rune‑carvers at work. | Hands‑on learning and chance to purchase unique items. | | Riverbank Amphitheater | Evening concerts, lantern shows. | Water reflections enhance the visual experience. | | Echo Tower | Access the Story‑Streaming network; submit your own tale. | Connects distant villages and preserves oral history. |
Held every full moon of the third season, the festival is the apex of Edrivan entertainment. Lira, as Grand Light‑Keeper, orchestrates a city‑wide symphony of lanterns, music, and performance.
Every Edrivan believes they are a creator. Whether it’s woven kelp tapestries, song‑spun poetry, or dance‑crafted stories, each citizen contributes to a communal tapestry of culture. Lira’s specialty is Light‑Weaving, an art where strands of bioluminescent algae are braided into intricate patterns that tell mythic narratives when illuminated.
This website is adapted from Nerfies, UniversalNER and LLaVA, licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. We thank the LLaMA team for giving us access to their models.
Usage and License Notices: The data abd code is intended and licensed for research use only. They are also restricted to uses that follow the license agreement of LLaMA, ChatGPT, and the original dataset used in the benchmark. The dataset is CC BY NC 4.0 (allowing only non-commercial use) and models trained using the dataset should not be used outside of research purposes.