PN
Preslav Nakov
cs.CLcs.AIcs.LGcs.IRcs.CEcs.SIcs.MAcs.SE
On Valency
published · living versionsW_b67cayr4·v1 · currentpublished
The Bottleneck Is Retrieval, Not Reading: Diagnosing Evidence-Based Generation with RAG for Long Context Question Answering
with Debopriyo Banerjee, Owais Aijaz, Ahmer Tabassum, Zhuohan Xie +1
1 version
W_dyvakna6·v1published
W_b54zs28u·v1published
Preprints & journals
45 papers in the corpus · 2025–2026Benchmarking Arabic Dialects for Math Problems2610.28607v1 · Noor Abo Mokh, Teresa Lynn, Nizar Habash et al.2026 · 0 citationsarXivon Valency
Aligning Cultural Behaviors in Arabic Language Models via Preferences2610.25362v1 · Junhong Liang, Daniil Orel, Saeed Almheiri et al.2026 · 0 citationsarXivon Valency
The Bottleneck Is Retrieval, Not Reading: Diagnosing Evidence-Based Generation with RAG for Long Context Question Answering2610.26951v1 · Debopriyo Banerjee, Owais Aijaz, Ahmer Tabassum et al.2026 · 0 citationsarXivon Valency
MedSNIP: Building and Benchmarking Snippet-Level Granularity for Medical Fact Verification2609.12884v1 · Hasan Iqbal, Sarfraz Ahmad, Hyunjae Kim et al.2026 · 0 citationsIn Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 11081-11104, Budapest, Hungary. Association for Computational Linguistics
Right Frame, Wrong Rule: Cultural Cues Expose the Financial Knowledge Gap They Were Meant to Close2609.00999v1 · Rania Elbadry, Ahmed Heakl, Saeed Almheiri et al.2026 · 0 citationsarXiv
EDRAC: Benchmarking Arabic Dialect Reading Comprehension2609.01113v1 · Noor Abo Mokh, Kirill Chirkunov, Teresa Lynn et al.2026 · 0 citationsarXiv
How Correct Is Your Answer? A Semantic Correctness Framework for Open QA Evaluation2609.01369v1 · Elitsa Yotkova, Violeta Kastreva, Petar Velkov et al.2026 · 0 citationsarXiv
Think with Structured Grounding: Perceptual Reinforcement Learning for Chart and Visual-Tabular Understanding2608.22429v1 · Changjiang Jiang, Qiannian Zhao, Lei Xin et al.2026 · 0 citationsarXiv
SurakshaEval: An Indic Safety Benchmark for Multilingual LLMs2608.07862v1 · Debopriyo Banerjee, Kapil Rajesh Kavitha, Angana Borah et al.2026 · 0 citationsarXiv
Seeing Is Not Deciding: Can Multimodal LLMs Act as Effective CEOs?2608.05864v1 · Yuyang Dai, Xueqing Peng, Yuxia Wang et al.2026 · 0 citationsarXiv
Overview of FinMMEval 2026 Task 1: Multilingual Financial Multiple-Choice Question Answering2607.19856v1 · Zhuohan Xie, Yuyang Dai, Rania Elbadry et al.2026 · 0 citationsarXiv
Overview of FinMMEval 2026 Task 2: Multilingual Financial Short-Answer Question Answering2607.19867v1 · Zhuohan Xie, Xueqing Peng, Georgi Georgiev et al.2026 · 0 citationsarXiv
Jais 2: A Family of Arabic-Centric Open Large Language Models2608.13580v1 · Mohamed Anwar, Abed Alhakim Freihat, George Ibrahim et al.2026 · 0 citationsarXiv
MIRAGE: Defending Long-Form RAG Against Misinformation Pollution2607.05069v1 · Saadeldine Eletter, Ruihong Zeng, Yuxia Wang et al.2026 · 0 citationsarXiv
Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs2607.03936v3 · Kareem Elozeiri, Mervat Abassy, Omar Kallas et al.2026 · 0 citationsarXiv
The FIL Hypothesis: Inductive Biases Help with Kernel Engineering2606.30442v1 · Nikolai Rozanov, Subhabrata Dutta, Preslav Nakov et al.2026 · 0 citationsarXiv
Bayesian control for coding agents2606.24453v1 · Theodore Papamarkou, Vladislav Smirnov, Viktor Mazanov et al.2026 · 0 citationsarXiv
ThinkBooster: A Unified Framework for Seamless Test-Time Scaling of LLM Reasoning2606.06915v2 · Vladislav Smirnov, Chieu Nguyen, Sergey Senichev et al.2026 · 0 citationsProceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 715-727, San Diego, California, USA, July 2026. Association for Computational Linguistics
UrduMMLU: A Massive Multitask Benchmark for Urdu Language Understanding2606.07167v2 · Ahmer Tabassum, Sarfraz Ahmad, Hasan Iqbal et al.2026 · 0 citationsIn Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 10341-10370, Budapest, Hungary. Association for Computational Linguistics
Better with Experience: Self-Evolving LLM Agents for Evidence-Grounded Health Community Notes2606.02215v2 · Zihang Fu, Fanxiao Li, Jianyang Gu et al.2026 · 0 citationsarXiv
Uncovering Temporal Framing in the News2606.00294v1 · Tarek Mahmoud, Veronika Solopova, Premtim Sahitaj et al.2026 · 1 citationarXiv
The Geometry of Forgetting: Temporal Knowledge Drift as an Independent Axis in LLM Representations2605.09195v1 · Rania Elbadry, Ahmed Heakl, Fan Zhang et al.2026 · 0 citationsarXiv
FMI_SU_Yotkova_Kastreva at SemEval-2026 Task 13: Lightweight Detection of LLM-Generated Code via Stylometric Signals2605.04157v1 · Elitsa Yotkova, Violeta Kastreva, Dimitar Dimitrov et al.2026 · 0 citationsarXiv
The Cylindrical Representation Hypothesis for Language Model Steering2605.01844v2 · Lang Gao, Jinghui Zhang, Wei Liu et al.2026 · 0 citationsarXiv
A Multi-View Media Profiling Suite: Resources, Evaluation, and Analysis2605.01336v1 · Muhammad Arslan Manzoor, Dilshod Azizov, Daniil Orel et al.2026 · 0 citationsarXiv
Instruction-Guided Poetry Generation in Arabic and Its Dialects2604.27766v1 · Abdelrahman Sadallah, Kareem Elozeiri, Mervat Abassy et al.2026 · 0 citationsarXiv
Cultural Benchmarking of LLMs in Standard and Dialectal Arabic Dialogues2605.00119v1 · Muhammad Dehan Al Kautsar, Saeed Almheiri, Momina Ahsan et al.2026 · 0 citationsarXiv
SAHM: A Benchmark for Arabic Financial and Shari'ah-Compliant Reasoning2604.19098v2 · Rania Elbadry, Sarfraz Ahmad, Ahmed Heakl et al.2026 · 1 citationACL 2026
Co-FactChecker: A Framework for Human-AI Collaborative Claim Verification Using Large Reasoning Models2604.13706v2 · Dhruv Sahnan, Subhabrata Dutta, Tanmoy Chakraborty et al.2026 · 0 citationsarXiv
Adaptive Conformal Prediction for Improving Factuality of Generations by Large Language Models2604.13991v1 · Aleksandr Rubashevskii, Dzianis Piatrashyn, Preslav Nakov et al.2026 · 0 citationsarXiv
Why Don't You Know? Evaluating the Impact of Uncertainty Sources on Uncertainty Quantification in LLMs2604.10495v2 · Maiya Goloburda, Roman Vashurin, Fedor Chernogorskii et al.2026 · 0 citationsarXiv
ReDAct: Uncertainty-Aware Deferral for LLM Agents2604.07036v1 · Dzianis Piatrashyn, Nikita Kotelevskii, Kirill Grishchenkov et al.2026 · 0 citationsarXiv
SciDER: Scientific Data-centric End-to-end Researcher2603.01421v3 · Ke Lin, Owais Aijaz, Yilin Lu et al.2026 · 0 citationsarXiv
DenoiseRank: Learning to Rank by Diffusion Models2604.20852v1 · Ying Wang, Preslav Nakov, Shangsong Liang2026 · 0 citationsarXiv
AICD Bench: A Challenging Benchmark for AI-Generated Code Detection2602.02079v1 · Daniil Orel, Dilshod Azizov, Indraneil Paul et al.2026 · 2 citationsarXiv
Health-ORSC-Bench: A Benchmark for Measuring Over-Refusal and Safety Completion in Health Context2601.17642v2 · Zhihao Zhang, Liting Huang, Guanghao Wu et al.2026 · 2 citationsarXiv
VISPA: Pluralistic Alignment via Automatic Value Selection and Activation2601.12758v2 · Shenyan Zheng, Jiayou Zhong, Anudeex Shetty et al.2026 · 0 citationsarXiv
MemeLens: Multilingual Multitask VLMs for Memes2601.12539v4 · Ali Ezzat Shahroor, Mohamed Bayan Kmainasi, Abul Hasnat et al.2026 · 0 citationsarXiv
YaPO: Learnable Sparse Activation Steering Vectors for Domain Adaptation2601.08441v1 · Abdelaziz Bounhar, Rania Hossam Elmohamady Elbadry, Hadi Abdine et al.2026 · 0 citationsarXiv
FinCARDS: Card-Based Analyst Reranking for Financial Document Question Answering2601.06992v2 · Yixi Zhou, Fan Zhang, Yu Chen et al.2026 · 1 citationarXiv
Beyond the Crowd: LLM-Augmented Community Notes for Governing Health Misinformation2510.11423v4 · Jiaying Wu, Zihang Fu, Haonan Wang et al.2025 · 0 citationsarXiv
Beyond Memorization: Extending Reasoning Depth with Recurrence, Memory and Test-Time Compute Scaling2508.16745v3 · Ivan Rodkin, Daniil Orel, Konstantin Smirnov et al.2025 · 0 citationsarXiv
UnsafeChain: Enhancing Reasoning Model Safety via Hard Cases2507.21652v2 · Raj Vardhan Tomar, Preslav Nakov, Yuxia Wang2025 · 1 citationarXiv
A Head to Predict and a Head to Question: Pre-trained Uncertainty Quantification Heads for Hallucination Detection in LLM Outputs2505.08200v1 · Artem Shelmanov, Ekaterina Fadeeva, Akim Tsvigun et al.2025 · 5 citationsarXiv
Can LLMs Automate Fact-Checking Article Writing?2503.17684v2 · Dhruv Sahnan, David Corney, Irene Larraz et al.2025 · 2 citationsarXiv
Career total: 56 works. 45 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.