Personalised text generation is essential for user-centric information systems, yet most evaluation methods overlook the individuality of users. We introduce \textbf{PREF}, a \textbf{P}ersonalised \textbf{R}eference-free \textbf{E}valuation \textbf{F}ramework that jointly measures general output quality and user-specific alignment without requiring gold personalised references. PREF operates in a three-step pipeline: (1) a coverage stage uses a large language model (LLM) to generate a comprehensive, query-specific guideline covering universal criteria such as factuality, coherence, and completeness; (2) a preference stage re-ranks and selectively augments these factors using the target user's profile, stated or inferred preferences, and context, producing a personalised evaluation rubric; and (3) a scoring stage applies an LLM judge to rate candidate answers against this rubric, ensuring baseline adequacy while capturing subjective priorities. This separation of coverage from preference improves robustness, transparency, and reusability, and allows smaller models to approximate the personalised quality of larger ones. Experiments on the PrefEval benchmark, including implicit preference-following tasks, show that PREF achieves higher accuracy, better calibration, and closer alignment with human judgments than strong baselines. By enabling scalable, interpretable, and user-aligned evaluation, PREF lays the groundwork for more reliable assessment and development of personalised language generation systems.

Large language models (LLMs) empowered by chain-of-thought reasoning have achieved impressive accuracy on complex tasks but suffer from excessive inference costs and latency when applied uniformly to all problems. We propose SABER (Switchable and Balanced Training for Efficient LLM Reasoning), a reinforcement learning framework that endows LLMs with user-controllable, token-budgeted reasoning. SABER first profiles each training example's base-model thinking token usage and assigns it to one of the predefined budget tiers. During fine-tuning, the model is guided by system prompts and length-aware rewards to respect its assigned budget. In parallel, we incorporate no-think examples to ensure the model remains reliable even when explicit reasoning is turned off. SABER further supports four discrete inference modes--NoThink, FastThink, CoreThink, and DeepThink, enabling flexible trade-offs between latency and reasoning depth. Extensive evaluations on math reasoning (MA TH, GSM8K), code generation (MBPP), and logical reasoning (LiveBench-Reasoning) demonstrate that SABER achieves high accuracy under tight budgets, graceful degradation, and effective cross-scale and cross-domain generalization. In particular, SABER-FastThink cuts reasoning length by 65.4% and yields a 3.6% accuracy gain compared with the base model on the MA TH benchmark.

Among the many challenges mothers undergo after childbirth, postpartum depression ( ppd) is a severe condition that significantly impacts their mental and physical well-being. Consequently, the rapid detection of ppd and their associated risk factors is critical for in-time assessment and intervention through specialized prevention procedures. Accordingly, this work addresses the need to help practitioners make decisions with the latest technological advancements to enable real-time screening and treatment recommendations. Mainly, our work contributes to an intelligent ppd screening system that combines Natural Language Processing, Machine Learning ( ml), and Large Language Models ( llm s) towards an affordable, real-time, and non-invasive free speech analysis. Moreover, it addresses the black box problem since the predictions are described to the end users thanks to the combination of llm s with interpretable ml models ( i.e., tree-based algorithms) using feature importance and natural language. The results obtained are 90 % on ppd detection for all evaluation metrics, outperforming the competing solutions in the literature. Ultimately, our solution contributes to the rapid detection of ppd and their associated risk factors, critical for in-time and proper assessment and intervention. Introduction Depression is a global public health concern that affects more than 150 million people, being more prevalent in women (Labaka et al., 2018; Moreira et al., 2019). Among the many challenges mothers undergo after childbirth, postpartum depression ( ppd) is a severe condition that usually requires medical intervention (Falana & Carrington, 2019). Mainly, ppd is a common non-psychotic mental disorder during the first year after childbirth that can lead to severe complications in the women's health (Abadiga, 2019). Current data indicates that between 10 % to 15 % of mothers worldwide are affected with ppd yearly (Fatima et al., 2019; Liu et al., 2023). Moreover, only 20% of the target population is diagnosed or even treated promptly (Mazumder & Baruah, 2021).

Test-Time Compute (TTC) has emerged as a powerful paradigm for enhancing the performance of Large Language Models (LLMs) at inference, leveraging strategies such as Test-Time Training (TTT) and Retrieval-Augmented Generation (RAG). However, the optimal adaptation strategy varies across queries, and indiscriminate application of TTC strategy incurs substantial computational overhead. In this work, we introduce Reward-Guided Test-Time Compute (RTTC), a novel framework that adaptively selects the most effective TTC strategy for each query via a pretrained reward model, maximizing downstream accuracy across diverse domains and tasks. RTTC operates in a distributed server-client architecture, retrieving relevant samples from a remote knowledge base and applying RAG or lightweight fine-tuning on client devices only when necessary. To further mitigate redundant computation, we propose Query-State Caching, which enables the efficient reuse of historical query states at both retrieval and adaptation levels. Extensive experiments across multiple LLMs and benchmarks demonstrate that RTTC consistently achieves superior accuracy compared to vanilla RAG or TTT, validating the necessity of adaptive, reward-guided TTC selection and the potential of RTTC for scalable, high-performance language model adaptation.

This study introduces a significance testing-enhanced conformal prediction (CP) framework to improve trustworthiness of large language models (LLMs) in multiple-choice question answering (MCQA). While LLMs have been increasingly deployed in disciplinary QA scenarios, hallucination and nonfactual generation substantially compromise response reliability. Although CP provides statistically rigorous marginal coverage guarantees for prediction sets, and significance testing offers established statistical rigor, their synergistic integration remains unexplored. To mitigate hallucination and factual inaccuracies, our framework integrates $p$-value computation with conformity scoring through self-consistency resampling of MCQA responses. This approach calculates option frequencies to address LLMs' black-box nature, subsequently constructing prediction sets via null hypothesis testing ($\mathcal{H}_0$) with empirically derived $p$-values. Evaluations on MMLU and MMLU-Pro benchmarks using off-the-shelf LLMs demonstrate: (1) The enhanced CP achieves user-specified empirical miscoverage rates; (2) Test-set average prediction set size (APSS) decreases monotonically with increasing risk levels ($α$), validating APSS as an effective uncertainty metric. This work establishes a principled statistical framework for trustworthy LLM deployment in high-stakes QA applications.

Efficiently enhancing the reasoning capabilities of large language models (LLMs) in federated learning environments remains challenging, particularly when balancing performance gains with strict computational, communication, and privacy constraints. This challenge is especially acute in healthcare, where decisions-spanning clinical, operational, and patient-facing contexts-demand not only accurate outputs but also interpretable, traceable rationales to ensure safety, accountability, and regulatory compliance. Conventional federated tuning approaches on LLM fail to address this need: they optimize primarily for answer correctness while neglecting rationale quality, leaving CoT capabilities dependent on models' innate pre-training abilities. Moreover, existing methods for improving rationales typically rely on privacy-violating knowledge distillation from centralized models. Additionally, the communication overhead in traditional federated fine-tuning on LLMs remains substantial. We addresses this gap by proposing FedCoT, a novel framework specifically designed to enhance reasoning in federated settings. FedCoT leverages a lightweight chain-of-thought enhancement mechanism: local models generate multiple reasoning paths, and a compact discriminator dynamically selects the most promising one. This approach improves reasoning accuracy and robustness while providing valuable interpretability, which is particularly critical for medical applications. To manage client heterogeneity efficiently, we adopt an improved aggregation approach building upon advanced LoRA module stacking, incorporating client classifier-awareness to achieve noise-free aggregation across diverse clients. Comprehensive experiments on medical reasoning tasks demonstrate that FedCoT significantly boosts client-side reasoning performance under stringent resource budgets while fully preserving data privacy.

Natural language is replete with superficially different statements, such as ``Charles Darwin wrote" and ``Charles Darwin is the author of", which carry the same meaning. Large language models (LLMs) should generate the same next-token probabilities in such cases, but usually do not. Empirical workarounds have been explored, such as using k-NN estimates of sentence similarity to produce smoothed estimates. In this paper, we tackle this problem more abstractly, introducing a categorical homotopy framework for LLMs. We introduce an LLM Markov category to represent probability distributions in language generated by an LLM, where the probability of a sentence, such as ``Charles Darwin wrote" is defined by an arrow in a Markov category. However, this approach runs into difficulties as language is full of equivalent rephrases, and each generates a non-isomorphic arrow in the LLM Markov category. To address this fundamental problem, we use categorical homotopy techniques to capture ``weak equivalences" in an LLM Markov category. We present a detailed overview of application of categorical homotopy to LLMs, from higher algebraic K-theory to model categories, building on powerful theoretical results developed over the past half a century.

In recent years, large language models (LLMs) have achieved remarkable advancements in multimodal processing, including end-to-end speech-based language models that enable natural interactions and perform specific tasks in task-oriented dialogue (TOD) systems. However, existing TOD datasets are predominantly text-based, lacking real speech signals that are essential for evaluating the robustness of speech-based LLMs. Moreover, existing speech TOD datasets are primarily English and lack critical aspects such as speech disfluencies and speaker variations. To address these gaps, we introduce RealTalk-CN, the first Chinese multi-turn, multi-domain speech-text dual-modal TOD dataset, comprising 5.4k dialogues (60K utterances, 150 hours) with paired speech-text annotations. RealTalk-CN captures diverse dialogue scenarios with annotated spontaneous speech disfluencies, ensuring comprehensive coverage of real-world complexities in speech dialogue. In addition, we propose a novel cross-modal chat task that authentically simulates real-world user interactions, allowing dynamic switching between speech and text modalities. Our evaluation covers robustness to speech disfluencies, sensitivity to speaker characteristics, and cross-domain performance. Extensive experiments validate the effectiveness of RealTalk-CN, establishing a strong foundation for Chinese speech-based LLMs research.

Current role-play studies often rely on unvalidated LLM-as-a-judge paradigms, which may fail to reflect how humans perceive role fidelity. A key prerequisite for human-aligned evaluation is role identification, the ability to recognize who is speaking based on dialogue context. We argue that any meaningful judgment of role-playing quality (how well a character is played) fundamentally depends on first correctly attributing words and actions to the correct persona (who is speaking). We present PersonaEval, the first benchmark designed to test whether LLM evaluators can reliably identify human roles. PersonaEval uses human-authored dialogues from novels, scripts, and video transcripts, challenging models to determine the correct persona according to the conversation context. Our experiments, including a human study, show that even the best-performing LLMs reach only around 69% accuracy, well below the level needed for reliable evaluation. In contrast, human participants perform near ceiling with 90.8% accuracy, highlighting that current LLM evaluators are still not human enough to effectively judge role-play scenarios. To better understand this gap, we examine training-time adaptation and test-time compute, suggesting that reliable evaluation requires more than task-specific tuning, but depends on strong, human-like reasoning abilities in LLM evaluators. We release our benchmark at https://github.com/maple-zhou/PersonaEval.

Large language model (LLM) training faces a critical bottleneck: the scarcity of high-quality, reasoning-intensive question-answer pairs, especially from sparse, domain-specific sources like PubMed papers or legal documents. Existing methods rely on surface patterns, fundamentally failing to generate controllable, complex multi-hop reasoning questions that test genuine understanding-essential for advancing LLM training paradigms. We present \textbf{Semantic Bridge}, the first universal framework for controllably generating sophisticated multi-hop reasoning questions from arbitrary sources. Our breakthrough innovation is \textit{semantic graph weaving}-three complementary bridging mechanisms (entity bridging for role-varying shared entities, predicate chain bridging for temporal/causal/logical sequences, and causal bridging for explicit reasoning chains)-that systematically construct complex pathways across documents, with fine-grained control over complexity and types via AMR-driven analysis. Our multi-modal AMR pipeline achieves up to 9.5% better round-trip quality, enabling production-ready controllable QA generation. Extensive evaluation demonstrates performance across both general-purpose datasets (Wikipedia) and specialized domains (biomedicine) It yields consistent 18.3%-25.4% gains over baselines across four languages (English, Chinese, French, German). Question pairs generated from 200 sources outperform 600 native human annotation examples with 67% fewer materials. Human evaluation shows 23.4% higher complexity, 18.7% better answerability, and 31.2% improved pattern coverage. Semantic Bridge establishes a new paradigm for LLM training data synthesis, enabling controllable generation of targeted reasoning questions from sparse sources. We will release our core code and semantic bridge model.