Amr Keleg

I realized the instructions were circulated as part of the acceptance email. This caught me by surprise as well, especially that it's different from how things were previously managed.

All the best with your preparations, though!

According to openreview, our paper got accepted to #ACL2025NLP 🥳🥳🥳
Further details to be shared soon!

Glad to know you liked it 😀

Congrats, Arij!
🎉🎉

I share some preliminary thoughts for four steps that could help in building culturally representative models.

Lastly, I hope this will spark discussions within the Arabic NLP community, and the broader NLP community interested in serving marginalized speech communities!

(4/4)

* The NLP community acknowledges the rich diversity of the Arabic dialects, which are a manifestation of cultural differences across the region.

* While Arabic-specific LLMs are still marketed as serving all Arabs, our alignment data/benchmarks are scarce and not inclusive enough!

(3/4)

* Arabic speakers have substantial cultural similarities (see map below). This does not imply they have one single homogenous culture!

* Their views tend to be ignored, even for largely diverse alignment datasets (e.g., PRISM, Kirk et al., 2024).

(2/4)

LLM Alignment for the Arabs: A Homogenous Culture or Diverse Ones? Large language models (LLMs) have the potential of being useful tools that can automate tasks and assist humans. However, these models are more fluent in English and more aligned with Western cultures, norms, and values. Arabic-specific LLMs are being developed to better capture the nuances of the Arabic language, as well as the views of the Arabs. Yet, Arabs are sometimes assumed to share the same culture. In this position paper, I discuss the limitations of this assumption and provide preliminary thoughts for how to build systems that can better represent the cultural diversity within the Arab world. The invalidity of the cultural homogeneity assumption might seem obvious, yet, it is widely adopted in developing multilingual and Arabic-specific LLMs. I hope that this paper will encourage the NLP community to be considerate of the cultural diversity within various communities speaking the same language.

My position paper “LLM Alignment for the Arabs: A Homogenous Culture or Diverse Ones?” got accepted to the @c3nlp.bsky.social workshop co-located with @naaclmeeting.bsky.social

I share concerns about the missed opportunities with the rise of Arabic-specific LLMs.

📄 arxiv.org/abs/2503.15003
(1/4)

I share some preliminary thoughts for four steps that could help in building culturally representative models.

Lastly, I hope this will spark discussions within the Arabic NLP community, and the broader NLP community interested in serving marginalized speech communities!

(4/4)

* The NLP community acknowledges the rich diversity of the Arabic dialects, which are a manifestation of cultural differences across the region.

* While Arabic-specific LLMs are still marketed as serving all Arabs, our alignment data/benchmarks are scarce and not inclusive enough!

(3/4)

* Arabic speakers have substantial cultural similarities (see map below). This does not imply they have one single homogenous culture!

* Their views tend to be ignored, even for largely diverse alignment datasets (e.g., PRISM, Kirk et al., 2024).

(2/4)