Kidist Amde Mekonnen
@kidist-amde.bsky.social
Ph.D. Candidate @ IRLab, UvA | AI Researcher | Information Retrieval | Generative AI | Passionate about responsible AI, multi-modal learning, and tech for real-world impact | Lifelong learner 🌱
🎉 Excited to share that our paper has been accepted to ACL 2025 Findings
“Optimized Text Embedding Models and Benchmarks for Amharic Passage Retrieval”
Check out Our Resources:
📄 lnkd.in/eN5m_gQV
💻 : lnkd.in/eHWFGE5t
🤗: lnkd.in/e4wjKNbV
“Optimized Text Embedding Models and Benchmarks for Amharic Passage Retrieval”
Check out Our Resources:
📄 lnkd.in/eN5m_gQV
💻 : lnkd.in/eHWFGE5t
🤗: lnkd.in/e4wjKNbV
LinkedIn
This link will take you to a page that’s not on LinkedIn
lnkd.in
May 27, 2025 at 12:05 PM
🎉 Excited to share that our paper has been accepted to ACL 2025 Findings
“Optimized Text Embedding Models and Benchmarks for Amharic Passage Retrieval”
Check out Our Resources:
📄 lnkd.in/eN5m_gQV
💻 : lnkd.in/eHWFGE5t
🤗: lnkd.in/e4wjKNbV
“Optimized Text Embedding Models and Benchmarks for Amharic Passage Retrieval”
Check out Our Resources:
📄 lnkd.in/eN5m_gQV
💻 : lnkd.in/eHWFGE5t
🤗: lnkd.in/e4wjKNbV
Excited to share our paper “Lightweight and Direct Document Relevance Optimization for Generative IR” is accepted as a full paper at #SIGIR2025🎉
We present DDRO: a GenIR training method that skips RL and directly aligns docid generation with relevance using a simple pairwise optimization step.
We present DDRO: a GenIR training method that skips RL and directly aligns docid generation with relevance using a simple pairwise optimization step.
April 9, 2025 at 1:00 PM
Excited to share our paper “Lightweight and Direct Document Relevance Optimization for Generative IR” is accepted as a full paper at #SIGIR2025🎉
We present DDRO: a GenIR training method that skips RL and directly aligns docid generation with relevance using a simple pairwise optimization step.
We present DDRO: a GenIR training method that skips RL and directly aligns docid generation with relevance using a simple pairwise optimization step.