Model merging aggregates Large Language Models (LLMs) finetuned on different tasks into a stronger one. However, parameter conflicts between models leads to performance degradation in averaging. While model routing addresses this issue by selecting individual models during inference, it imposes excessive storage and compute costs, and fails to leverage the common knowledge from different models. In this work, we observe that different layers exhibit varying levels of parameter conflicts. Building on this insight, we average layers with minimal parameter conflicts and use a novel task-level expert routing for layers with significant conflicts. To further reduce storage costs, inspired by task arithmetic sparsity, we decouple multiple fine-tuned experts into a dense expert and several sparse experts. Considering the out-of-distribution samples, we select and merge appropriate experts based on the task uncertainty of the input data. We conduct extensive experiments on both LLaMA and Qwen with varying parameter scales, and evaluate on real-world reasoning tasks. Results demonstrate that our method consistently achieves significant performance improvements while requiring less system cost compared to existing methods.

This study presents a framework for automated evaluation of dynamically evolving topic taxonomies in scientific literature using Large Language Models (LLMs). In digital library systems, topic modeling plays a crucial role in efficiently organizing and retrieving scholarly content, guiding researchers through complex knowledge landscapes. As research domains proliferate and shift, traditional human centric and static evaluation methods struggle to maintain relevance. The proposed approach harnesses LLMs to measure key quality dimensions, such as coherence, repetitiveness, diversity, and topic-document alignment, without heavy reliance on expert annotators or narrow statistical metrics. Tailored prompts guide LLM assessments, ensuring consistent and interpretable evaluations across various datasets and modeling techniques. Experiments on benchmark corpora demonstrate the method's robustness, scalability, and adaptability, underscoring its value as a more holistic and dynamic alternative to conventional evaluation strategies.

Large language models (LLMs) show promise for health applications when combined with behavioral sensing data. Traditional approaches convert sensor data into text prompts, but this process is prone to errors, computationally expensive, and requires domain expertise. These challenges are particularly acute when processing extended time series data. While time series foundation models (TFMs) have recently emerged as powerful tools for learning representations from temporal data, bridging TFMs and LLMs remains challenging. Here, we present Time2Lang, a framework that directly maps TFM outputs to LLM representations without intermediate text conversion. Our approach first trains on synthetic data using periodicity prediction as a pretext task, followed by evaluation on mental health classification tasks. We validate Time2Lang on two longitudinal wearable and mobile sensing datasets: daily depression prediction using step count data (17,251 days from 256 participants) and flourishing classification based on conversation duration (46 participants over 10 weeks). Time2Lang maintains near constant inference times regardless of input length, unlike traditional prompting methods. The generated embeddings preserve essential time-series characteristics such as auto-correlation. Our results demonstrate that TFMs and LLMs can be effectively integrated while minimizing information loss and enabling performance transfer across these distinct modeling paradigms. To our knowledge, we are the first to integrate a TFM and an LLM for health, thus establishing a foundation for future research combining general-purpose large models for complex healthcare tasks.

We present a novel data set, WhoDunIt, to assess the deductive reasoning capabilities of large language models (LLM) within narrative contexts. Constructed from open domain mystery novels and short stories, the dataset challenges LLMs to identify the perpetrator after reading and comprehending the story. To evaluate model robustness, we apply a range of character-level name augmentations, including original names, name swaps, and substitutions with well-known real and/or fictional entities from popular discourse. We further use various prompting styles to investigate the influence of prompting on deductive reasoning accuracy. We conduct evaluation study with state-of-the-art models, specifically GPT-4o, GPT-4-turbo, and GPT-4o-mini, evaluated through multiple trials with majority response selection to ensure reliability. The results demonstrate that while LLMs perform reliably on unaltered texts, accuracy diminishes with certain name substitutions, particularly those with wide recognition. This dataset is publicly available here.

Web browsing agents powered by large language models (LLMs) have shown tremendous potential in automating complex web-based tasks. Existing approaches typically rely on large LLMs (e.g., GPT-4o) to explore web environments and generate trajectory data, which is then used either for demonstration retrieval (for large LLMs) or to distill small LLMs (e.g., Llama3) in a process that remains decoupled from the exploration. In this paper, we propose AgentSymbiotic, an iterative framework that couples data synthesis with task-performance, yielding a "symbiotic improvement" for both large and small LLMs. Our study uncovers a complementary dynamic between LLM types: while large LLMs excel at generating high-quality trajectories for distillation, the distilled small LLMs-owing to their distinct reasoning capabilities-often choose actions that diverge from those of their larger counterparts. This divergence drives the exploration of novel trajectories, thereby enriching the synthesized data. However, we also observe that the performance of small LLMs becomes a bottleneck in this iterative enhancement process. To address this, we propose two innovations in LLM distillation: a speculative data synthesis strategy that mitigates off-policy bias, and a multi-task learning approach designed to boost the reasoning capabilities of the student LLM. Furthermore, we introduce a Hybrid Mode for Privacy Preservation to address user privacy concerns. Evaluated on the WEBARENA benchmark, AgentSymbiotic achieves SOTA performance with both LLM types. Our best Large LLM agent reaches 52%, surpassing the previous best of 45%, while our 8B distilled model demonstrates a competitive 49%, exceeding the prior best of 28%. Code will be released upon acceptance.

Background: Data collected in controlled settings typically results in high-quality datasets. However, in real-world applications, the quality of data collection is often compromised. It is well established that the quality of a dataset significantly impacts the performance of machine learning models. Methods: A rudimentary error rate metric was developed to evaluate textual dataset quality at the token level. Mixtral Large Language Model (LLM) was used to quantify and correct errors in low quality datasets. The study analyzed two healthcare datasets: the high-quality MIMIC-III public hospital dataset and a lower-quality private dataset from Australian aged care homes. Errors were systematically introduced into MIMIC at varying rates, while the ACH dataset quality was improved using the LLM. Results: For the sampled 35,774 and 6,336 patients from the MIMIC and ACH datasets respectively, we used Mixtral to introduce errors in MIMIC and correct errors in ACH. Mixtral correctly detected errors in 63% of progress notes, with 17% containing a single token misclassified due to medical terminology. LLMs demonstrated potential for improving progress note quality by addressing various errors. Under varying error rates, feature representation performance was tolerant to lower error rates (<10%) but declined significantly at higher rates. Conclusions: The study revealed that models performed relatively well on datasets with lower error rates (<10%), but their performance declined significantly as error rates increased (>=10%). Therefore, it is crucial to evaluate the quality of a dataset before utilizing it for machine learning tasks. For datasets with higher error rates, implementing corrective measures is essential to ensure the reliability and effectiveness of machine learning models.

In scientific research, the ability to effectively retrieve relevant documents based on complex, multifaceted queries is critical. Existing evaluation datasets for this task are limited, primarily due to the high costs and effort required to annotate resources that effectively represent complex queries. We developed a benchmark dataset within the field of computer science, consisting of 100 human-authored complex query cases. For each complex query, we assembled a collection of 100 relevant documents and produced annotated relevance scores for ranking them. Recognizing the significant labor of expert annotation, we also introduce Anno-GPT, a scalable framework for evaluating the viability of Large Language Models (LLMs) such as ChatGPT-3.5 for expert-level dataset annotation tasks.

A consortium led by Elon Musk said it has offered 97.4bn to buy the nonprofit that controls OpenAI, months after the billionaire sued the artificial intelligence startup to block it from transitioning to a for-profit firm. Musk's bid, revealed on Monday, could ratchet up longstanding tensions with OpenAI CEO Sam Altman over the future of the startup at the heart of a boom in generative AI technology. Altman promptly posted on X: "No thank you but we will buy twitter for 9.74 billion if you want." The two are already embroiled in an ongoing lawsuit. Musk criticised a 500bn OpenAI-led project called Stargate announced with great fanfare at the White House just after United States President Donald Trump returned to office, suggesting the investors involved lacked the funding for the project.

Elon Musk escalated his feud with OpenAI and its CEO Sam Altman on Monday. The billionaire is leading a consortium of investors that announced it had submitted a bid of 97.4bn for "all assets" of the artificial intelligence company to OpenAI's board of directors. The startup, which operates ChatGPT, has been working to restructure itself away from its original non-profit status. OpenAI also operates a for-profit subsidiary, and Musk's unsolicited offer could complicate the company's plans. The Wall Street Journal first reported the proposed bid. "If Sam Altman and the present OpenAI, Inc. Board of Directors are intent on becoming a fully for-profit corporation, it is vital that the charity be fairly compensated for what its leadership is taking away from it: control over the most transformative technology of our time," said Marc Toberoff, the attorney representing the investors.

OpenAI is widely credited with helping bring artificial intelligence tools into the mainstream and sparking huge investment in the sector. Musk and Altman co-founded the start-up in 2015 as a non-profit company, but the relationship has soured since the Tesla and X boss departed the firm in 2018. Altman is said to be restructuring the company to become a for-profit entity, stripping it of its non-profit board - a move Musk argues means the company has abandoned its founding mission of developing AI for the benefit of humanity. But OpenAI argues its transition into a for-profit firm is required to secure the money needed for developing the best artificial intelligence models. "It's time for OpenAI to return to the open-source, safety-focused force for good it once was. We will make sure that happens," Musk said in a statement.