As a key to accessing research impact, citation dynamics underpins research evaluation, scholarly recommendation, and the study of knowledge diffusion. Citation prediction is particularly critical for newborn papers, where early assessment must be performed without citation signals and under highly long-tailed distributions. We identify two key research gaps: (i) insufficient modeling of implicit factors of scientific impact, leading to reliance on coarse proxies; and (ii) a lack of bias-aware learning that can deliver stable predictions on lowly cited papers. We address these gaps by proposing a Bias-Aware Citation Prediction Framework, which combines multi-agent feature extraction with robust graph representation learning. First, a multi-agent x graph co-learning module derives fine-grained, interpretable signals, such as reproducibility, collaboration network, and text quality, from metadata and external resources, and fuses them with heterogeneous-network embeddings to provide rich supervision even in the absence of early citation signals. Second, we incorporate a set of robust mechanisms: a two-stage forward process that routes explicit factors through an intermediate exposure estimate, GroupDRO to optimize worst-case group risk across environments, and a regularization head that performs what-if analyses on controllable factors under monotonicity and smoothness constraints. Comprehensive experiments on two real-world datasets demonstrate the effectiveness of our proposed model. Specifically, our model achieves around a 13% reduction in error metrics (MALE and RMSLE) and a notable 5.5% improvement in the ranking metric (NDCG) over the baseline methods.

An interesting behavior in large language models (LLMs) is prompt sensitivity. When provided with different but semantically equivalent versions of the same prompt, models may produce very different distributions of answers. This suggests that the uncertainty reflected in a model's output distribution for one prompt may not reflect the model's uncertainty about the meaning of the prompt. We model prompt sensitivity as a type of generalization error, and show that sampling across the semantic concept space with paraphrasing perturbations improves uncertainty calibration without compromising accuracy. Additionally, we introduce a new metric for uncertainty decomposition in black-box LLMs that improves upon entropy-based decomposition by modeling semantic continuities in natural language generation. We show that this decomposition metric can be used to quantify how much LLM uncertainty is attributed to prompt sensitivity. Our work introduces a new way to improve uncertainty calibration in prompt-sensitive language models, and provides evidence that some LLMs fail to exhibit consistent general reasoning about the meanings of their inputs.

Advancements in Large Language Models (LLMs) and their increasing use in medical question-answering necessitate rigorous evaluation of their reliability.A critical challenge lies in hallucination, where models generate plausible yet factually incorrect outputs.In the medical domain, this poses serious risks to patient safety and clinical decision-making.To address this, we introduce MedHallu, one of the first benchmark specifically designed for medical hallucination detection.Med-Hallu comprises 10,000 high-quality questionanswer pairs derived from PubMedQA, with hallucinated answers systematically generated through a controlled pipeline.Our experiments show that state-of-the-art LLMs, including GPT-4o, Llama-3.1, and the medically fine-tuned UltraMedical, struggle with this binary hallucination detection task, with the best model achieving an F1 score as low as 0.625 for detecting "hard" category hallucinations.Using bidirectional entailment clustering, we show that harder-to-detect hallucinations are semantically closer to ground truth.Through experiments, we also show incorporating domainspecific knowledge and introducing a "not sure" category as one of the answer categories improves the precision and F1 scores by up to 38% relative to baselines.

Abstract While facilitating science, interdisciplinary research (IDR) has a heavier cognitive burden for researchers compared to unidisciplinary research (UDR). Yet, little has been known about patterns of knowledge integration and diffusion structures of IDR. Here we adopt a causal inference strategy, namely propensity score matching, with all journal publications in 2005 in Microsoft Academic Graph to better understand the IDR effect in various research fields. We use the diversity of reference fields of one paper as the proxy of the paper's interdisciplinarity and estimate the effect of a research article being IDR on its knowledge integration and diffusion measured by its high‐order citation/reference cascade. We find that, in disciplines where IDR articles are less popular, such as mathematics, physics, and chemistry, IDR needs a more extensive knowledge base than UDR to gain a similar number of citations. In disciplines where IDR articles are more popular, for example, psychology, geology, biology, and economics, a small knowledge base is enough for a high‐impact IDR article. As to knowledge diffusion, no matter whether IDR or UDR, a more extensive knowledge base leads to stronger knowledge diffusion ability. Findings imply potential drawbacks of pure interdisciplinarity‐oriented research policy; rather, the establishment of policies may vary across disciplines.