Large language models (LLMs) are increasingly used for both open-ended and structured tasks, yet their inference-time behavior is still largely dictated by heuristic decoding strategies such as greedy search, sampling, or reranking. These methods provide limited control and do not explicitly optimize for task-specific objectives. We introduce MEMETRON, a task-agnostic framework that formulates LLM decoding as a discrete black-box optimization problem. MEMETRON leverages hybrid metaheuristic algorithms, GENETRON and ANNETRON, to search the response space, guided by reward models and contextual operations performed by the LLM itself. This approach enables efficient discovery of high-reward responses without requiring model retraining or gradient access. The framework is modular and generalizes across diverse tasks, requiring only a reward function and lightweight prompt templates. We evaluate our framework on the critical human preference alignment task and demonstrate that it significantly outperforms standard decoding and reranking methods, highlighting its potential to improve alignment without model retraining.

Recent studies empirically reveal that large reasoning models (LRMs) can automatically allocate more reasoning strengths (i.e., the number of reasoning tokens) for harder problems, exhibiting difficulty-awareness for better task performance. While this automatic reasoning strength allocation phenomenon has been widely observed, its underlying mechanism remains largely unexplored. To this end, we provide explanations for this phenomenon from the perspective of model activations. We find evidence that LRMs pre-plan the reasoning strengths in their activations even before generation, with this reasoning strength causally controlled by the magnitude of a pre-allocated directional vector. Specifically, we show that the number of reasoning tokens is predictable solely based on the question activations using linear probes, indicating that LRMs estimate the required reasoning strength in advance. We then uncover that LRMs encode this reasoning strength through a pre-allocated directional vector embedded in the activations of the model, where the vector's magnitude modulates the reasoning strength. Subtracting this vector can lead to reduced reasoning token number and performance, while adding this vector can lead to increased reasoning token number and even improved performance. We further reveal that this direction vector consistently yields positive reasoning length prediction, and it modifies the logits of end-of-reasoning token to affect the reasoning length. Finally, we demonstrate two potential applications of our findings: overthinking behavior detection and enabling efficient reasoning on simple problems. Our work provides new insights into the internal mechanisms of reasoning in LRMs and offers practical tools for controlling their reasoning behaviors. Our code is available at https://github.com/AlphaLab-USTC/LRM-plans-CoT.

Large language models (LLMs) have demonstrated remarkable performance on various medical benchmarks, but their capabilities across different cognitive levels remain underexplored. Inspired by Bloom's Taxonomy, we propose a multi-cognitive-level evaluation framework for assessing LLMs in the medical domain in this study. The framework integrates existing medical datasets and introduces tasks targeting three cognitive levels: preliminary knowledge grasp, comprehensive knowledge application, and scenario-based problem solving. Using this framework, we systematically evaluate state-of-the-art general and medical LLMs from six prominent families: Llama, Qwen, Gemma, Phi, GPT, and DeepSeek. Our findings reveal a significant performance decline as cognitive complexity increases across evaluated models, with model size playing a more critical role in performance at higher cognitive levels. Our study highlights the need to enhance LLMs' medical capabilities at higher cognitive levels and provides insights for developing LLMs suited to real-world medical applications.

Recent voxel-wise multimodal brain encoding studies have shown that multimodal large language models (MLLMs) exhibit a higher degree of brain alignment compared to unimodal models in both unimodal and multimodal stimulus settings. More recently, instruction-tuned multimodal models have shown to generate task-specific representations that align strongly with brain activity. However, prior work evaluating the brain alignment of MLLMs has primarily focused on unimodal settings or relied on non-instruction-tuned multimodal models for multimodal stimuli. To address this gap, we investigated brain alignment, that is, measuring the degree of predictivity of neural activity recorded while participants were watching naturalistic movies (video along with audio) with representations derived from MLLMs. We utilized instruction-specific embeddings from six video and two audio instruction-tuned MLLMs. Experiments with 13 video task-specific instructions show that instruction-tuned video MLLMs significantly outperform non-instruction-tuned multimodal (by 15%) and unimodal models (by 20%). Our evaluation of MLLMs for both video and audio tasks using language-guided instructions shows clear disentanglement in task-specific representations from MLLMs, leading to precise differentiation of multimodal functional processing in the brain. We also find that MLLM layers align hierarchically with the brain, with early sensory areas showing strong alignment with early layers, while higher-level visual and language regions align more with middle to late layers. These findings provide clear evidence for the role of task-specific instructions in improving the alignment between brain activity and MLLMs, and open new avenues for mapping joint information processing in both the systems. We make the code publicly available [https://github.com/subbareddy248/mllm_videos].

Large language models (LLMs) are increasingly used to extract clinical data from electronic health records (EHRs), offering significant improvements in scalability and efficiency for real-world data (RWD) curation in oncology. However, the adoption of LLMs introduces new challenges in ensuring the reliability, accuracy, and fairness of extracted data, which are essential for research, regulatory, and clinical applications. Existing quality assurance frameworks for RWD and artificial intelligence do not fully address the unique error modes and complexities associated with LLM-extracted data. In this paper, we propose a comprehensive framework for evaluating the quality of clinical data extracted by LLMs. The framework integrates variable-level performance benchmarking against expert human abstraction, automated verification checks for internal consistency and plausibility, and replication analyses comparing LLM-extracted data to human-abstracted datasets or external standards. This multidimensional approach enables the identification of variables most in need of improvement, systematic detection of latent errors, and confirmation of dataset fitness-for-purpose in real-world research. Additionally, the framework supports bias assessment by stratifying metrics across demographic subgroups. By providing a rigorous and transparent method for assessing LLM-extracted RWD, this framework advances industry standards and supports the trustworthy use of AI-powered evidence generation in oncology research and practice.

Automated story writing has been a subject of study for over 60 years. Large language models can generate narratively consistent and linguistically coherent short fiction texts. Despite these advancements, rigorous assessment of such outputs for literary merit - especially concerning aesthetic qualities - has received scant attention. In this paper, we address the challenge of evaluating AI-generated microfictions and argue that this task requires consideration of literary criteria across various aspects of the text, such as thematic coherence, textual clarity, interpretive depth, and aesthetic quality. To facilitate this, we present GrAImes: an evaluation protocol grounded in literary theory, specifically drawing from a literary perspective, to offer an objective framework for assessing AI-generated microfiction. Furthermore, we report the results of our validation of the evaluation protocol, as answered by both literature experts and literary enthusiasts. This protocol will serve as a foundation for evaluating automatically generated microfictions and assessing their literary value.

The funeral director said "AI" as if it were a normal element of memorial services, like caskets or flowers. Of all places, I had not expected artificial intelligence to follow me into the small, windowless room of the mortuary. But here it was, ready to assist me in the task of making sense of death. It was already Wednesday, and I'd just learned that I had to write an obituary for my mother by Thursday afternoon if I wanted it to run in Sunday's paper. AI could help me do this.

Authors and publishers have filed multiple lawsuits over this issue, and in a new twist, researchers have shown that at least one AI model has not only used popular books in its training data, but also memorised their contents verbatim. But now, researchers have tested multiple models to see how much of that training data they can spit back out verbatim. They found that many models do not retain the exact text of the books in their training data – but one of Meta's models has memorised almost the entirety of certain books. If judges rule against the company, the researchers estimate that this could make Meta liable for at least 1 billion in damages. "That means, on the one hand, that AI models are not just'plagiarism machines', as some have alleged, but it also means that they do more than just learn general relationships between words," says Mark Lemley at Stanford University in California.

ChatGPT has been hit by a worldwide outage, sparking chaos in the corporate world. The AI has gained popularity in the workforce, helping employees draft the perfect email, research information and provide customer support. Students have also been left in the dark as they harness the intelligence to take exams and craft reports. 'We are observing elevated error rates and latency across ChatGPT and the API,' OpenAI shared on its sit. 'Our engineers have identified the root cause and are working as fast as possible to fix the issue.'

BleepingComputer reports that AI company OpenAI has suffered a major outage today affecting several of the company's AI services, including AI chatbot ChatGPT and AI video generator Sora. Though you can still access ChatGPT, it currently takes an unusually long time to respond and may end up responding with error messages. Similar issues with error rates and increased latency are affecting Sora and OpenAI's API, which could affect third-party services. According to Downdetector, thousands of users have been reporting outages, and the rate of reports hasn't slowed as of this writing. According to OpenAI's status tracker, the problem was first acknowledged at 2:36 AM this morning, and the issue is now marked as "Identified" with the company "still working on implementing the mitigation for this issue."