Marcus Engsig at DERC has developed DomiRank, a new centrality metric to quantify the dominance of nodes within networks. DomiRank integrates local and global topological information to determine the importance of each node for network stability. The research demonstrates that nodes with high DomiRank values indicate vulnerable areas heavily dependent on dominant nodes. Why it matters: This metric can help identify critical infrastructure components and vulnerabilities in complex systems, enhancing resilience against targeted attacks.
A new methodology emulating fact-checker criteria assesses news outlet factuality and bias using LLMs. The approach uses prompts based on fact-checking criteria to elicit and aggregate LLM responses for predictions. Experiments demonstrate improvements over baselines, with error analysis on media popularity and region, and a released dataset/code at https://github.com/mbzuai-nlp/llm-media-profiling.
Researchers from MBZUAI developed "uncertainty quantification heads" (UQ heads) to detect hallucinations in language models by probing internal states and estimating the credibility of generated text. UQ heads leverage attention maps and logits to identify potential hallucinations without altering the model's generation process or relying on external knowledge. The team found that UQ heads achieved state-of-the-art performance in claim-level hallucination detection across different domains and languages. Why it matters: This approach offers a more efficient and accurate method for identifying hallucinations, improving the reliability and trustworthiness of language models in various applications.
The paper introduces AraTrust, a new benchmark for evaluating the trustworthiness of LLMs when prompted in Arabic. The benchmark contains 522 multiple-choice questions covering dimensions like truthfulness, ethics, safety, and fairness. Experiments using AraTrust showed that GPT-4 performed the best, while open-source models like AceGPT 7B and Jais 13B had lower scores. Why it matters: This benchmark addresses a critical gap in evaluating LLMs for Arabic, which is essential for ensuring the safe and ethical deployment of AI in the Arab world.
A new paper coauthored by researchers at The University of Melbourne and MBZUAI explores disagreement in human annotation for AI training. The paper treats disagreement as a signal (human label variation or HLV) rather than noise, and proposes new evaluation metrics based on fuzzy set theory. These metrics adapt accuracy and F-score to cases where multiple labels may plausibly apply, aligning model output with the distribution of human judgments. Why it matters: This research addresses a key challenge in NLP by accounting for the inherent ambiguity in human language, potentially leading to more robust and human-aligned AI systems.
A new framework for constructing confidence sets for causal orderings within structural equation models (SEMs) is presented. It leverages a residual bootstrap procedure to test the goodness-of-fit of causal orderings, quantifying uncertainty in causal discovery. The method is computationally efficient and suitable for medium-sized problems while maintaining theoretical guarantees as the number of variables increases. Why it matters: This offers a new dimension of uncertainty quantification that enhances the robustness and reliability of causal inference in complex systems, but there is no indication of connection to the Middle East.
MBZUAI researchers have developed 'Byzantine antidote' (Bant), a novel defense mechanism against Byzantine attacks in federated learning. Bant uses trust scores and a trial function to dynamically filter and neutralize corrupted updates, even when a majority of nodes are compromised. The research was presented at the 40th Annual AAAI Conference on Artificial Intelligence.
Dr. Xinwei Sun from Microsoft Research Asia presented research on trustworthy AI, focusing on statistical learning with theoretical guarantees. The work covers methods for sparse recovery with false-discovery rate analysis and causal inference tools for robustness and explainability. Consistency and identifiability were addressed theoretically, with applications shown in medical imaging analysis. Why it matters: The research contributes to addressing key limitations of current AI models regarding explainability, reproducibility, robustness, and fairness, which are crucial for real-world applications in sensitive fields like healthcare.