With the rapid growth of smart home IoT devices, users are increasingly exposed to various security risks, as evident from recent studies. While seeking answers to know more on those security concerns, users are mostly left with their own discretion while going through various sources, such as online blogs and technical manuals, which may render higher complexity to regular users trying to extract the necessary information. This requirement does not go along with the common mindsets of smart home users and hence threatens the security of smart homes furthermore. In this paper, we aim to identify and address the major user-level security concerns in smart homes. Specifically, we develop a novel dataset of Q&A from public forums, capturing practical security challenges faced by smart home users. We extract major security concerns in smart homes from our dataset by leveraging the Latent Dirichlet Allocation (LDA). We fine-tune relatively "smaller" transformer models, such as T5 and Flan-T5, on this dataset to build a QA system tailored for smart home security. Unlike larger models like GPT and Gemini, which are powerful but often resource hungry and require data sharing, smaller models are more feasible for deployment in resource-constrained or privacy-sensitive environments like smart homes. The dataset is manually curated and supplemented with synthetic data to explore its potential impact on model performance. This approach significantly improves the system's ability to deliver accurate and relevant answers, helping users address common security concerns with smart home IoT devices. Our experiments on real-world user concerns show that our work improves the performance of the base models.

The Internet Yellow Pages (IYP) aggregates information from multiple sources about Internet routing into a unified, graph-based knowledge base. However, querying it requires knowledge of the Cypher language and the exact IYP schema, thus limiting usability for non-experts. In this paper, we propose ChatIYP, a domain-specific Retrieval-Augmented Generation (RAG) system that enables users to query IYP through natural language questions. Our evaluation demonstrates solid performance on simple queries, as well as directions for improvement, and provides insights for selecting evaluation metrics that are better fit for IYP querying AI agents.

We critically assess the efficacy of the current SOTA in hallucination detection and find that its performance on the RAGTruth dataset is largely driven by a spurious correlation with data. Controlling for this effect, state-of-the-art performs no better than supervised linear probes, while requiring extensive hyperparameter tuning across datasets. Out-of-distribution generalization is currently out of reach, with all of the analyzed methods performing close to random. We propose a set of guidelines for hallucination detection and its evaluation.

Large language models (LLMs) have attracted significant attention due to their impressive general capabilities across diverse downstream tasks. However, without domain-specific optimization, they often underperform on specialized knowledge benchmarks and even produce hallucination. Recent studies show that strategically infusing domain knowledge during pretraining can substantially improve downstream performance. A critical challenge lies in balancing this infusion trade-off: injecting too little domain-specific data yields insufficient specialization, whereas excessive infusion triggers catastrophic forgetting of previously acquired knowledge. In this work, we focus on the phenomenon of memory collapse induced by over-infusion. Through systematic experiments, we make two key observations, i.e. 1) Critical collapse point: each model exhibits a threshold beyond which its knowledge retention capabilities sharply degrade. 2) Scale correlation: these collapse points scale consistently with the model's size. Building on these insights, we propose a knowledge infusion scaling law that predicts the optimal amount of domain knowledge to inject into large LLMs by analyzing their smaller counterparts. Extensive experiments across different model sizes and pertaining token budgets validate both the effectiveness and generalizability of our scaling law.

As academic paper publication numbers grow exponentially, conducting in-depth surveys with LLMs automatically has become an inevitable trend. Outline writing, which aims to systematically organize related works, is critical for automated survey generation. Yet existing automatic survey methods treat outline writing as mere workflow steps in the overall pipeline. Such template-based workflows produce outlines that lack in-depth understanding of the survey topic and fine-grained styles. To address these limitations, we propose Meow, the first metadata-driven outline writing framework that produces organized and faithful outlines efficiently. Specifically, we first formulate outline writing as an end-to-end task that generates hierarchical structured outlines from paper metadata. We then curate a high-quality dataset of surveys from arXiv, bioRxiv, and medRxiv, and establish systematic evaluation metrics for outline quality assessment. Finally, we employ a two-stage training approach combining supervised fine-tuning and reinforcement learning. Our 8B reasoning model demonstrates strong performance with high structural fidelity and stylistic coherence.

We propose a small-scale language model (SLM) based agent architecture, Planner-Caller-Generator (P-C-G), optimized for Korean tool use. P-C-G separates planning, calling, and generation by role: the Planner produces an initial batch plan with limited on-demand replanning; the Caller returns a normalized call object after joint schema-value validation; and the Generator integrates tool outputs to produce the final answer. We apply a Korean-first value policy to reduce execution failures caused by frequent Korean-to-English code switching in Korean settings. Evaluation assumes Korean queries and Korean tool/parameter specifications; it covers single-chain, multi-chain, missing-parameters, and missing-functions scenarios, and is conducted via an LLM-as-a-Judge protocol averaged over five runs under a unified I/O interface. Results show that P-C-G delivers competitive tool-use accuracy and end-to-end quality while reducing tokens and maintaining acceptable latency, indicating that role-specialized SLMs are a cost-effective alternative for Korean tool-use agents.

Large language models (LLMs) deliver impressive generation quality, but incur very high inference cost because each output token is generated auto-regressively through all model layers. Early-exit based self-speculative decoding (EESD) has emerged to mitigate this cost. However, in practice, many approaches struggle to achieve the expected acceleration in such draft-then-verify paradigm even with a well-aligned early-exit head and selected exit position. Our analysis reveals that EESD only pays off when the vast majority of draft tokens are accepted by the LLM. Otherwise, the draft cost may overcome the acceleration gain and lead to a negative speedup. To mitigate this, we propose Pipeline-Parallel Self-Speculative Decoding (PPSD) that fully pipelines the draft and verification work so that no effort is wasted on failed predictions. It has two key innovations. We configure the model layers as a pipeline in which early-exit (draft) computations and remaining-layer (verification) computations overlap. We interleave drafting and verification per token. While the LLM is verifying the current token in its final layers, the early-exit path simultaneously drafts the next token. Such a verify-while-draft scheme keeps all units busy and validates tokens on-the-fly analogous to pipelining the speculation and verification stages. Empirical results confirm that PPSD achieves state-of-the-art acceleration in self-speculative LLM inference. On diverse benchmarks, PPSD achieves speedup ratios in the range of 2.01x~3.81x, which gains almost the optimal acceleration at the fixed acceptance rate and exit position, showcasing its advancement in providing efficient self-speculation.

The BERTopic framework leverages transformer embeddings and hierarchical clustering to extract latent topics from unstructured text corpora. While effective, it often struggles with social media data, which tends to be noisy and sparse, resulting in an excessive number of overlapping topics. Recent work explored the use of large language models for end-to-end topic modelling. However, these approaches typically require significant computational overhead, limiting their scalability in big data contexts. In this work, we propose a framework that combines BERTopic for topic generation with large language models for topic reduction. The method first generates an initial set of topics and constructs a representation for each. These representations are then provided as input to the language model, which iteratively identifies and merges semantically similar topics. We evaluate the approach across three Twitter/X datasets and four different language models. Our method outperforms the baseline approach in enhancing topic diversity and, in many cases, coherence, with some sensitivity to dataset characteristics and initial parameter selection.

Standard offline evaluations for language models -- a series of independent, state-less inferences made by models -- fail to capture how language models actually behave in practice, where personalization fundamentally alters model behavior. For instance, identical benchmark questions to the same language model can produce markedly different responses when prompted to a state-less system, in one user's chat session, or in a different user's chat session. In this work, we provide empirical evidence showcasing this phenomenon by comparing offline evaluations to field evaluations conducted by having 800 real users of ChatGPT and Gemini pose benchmark and other provided questions to their chat interfaces.

Recent years have witnessed a growing interest in personalizing the responses of large language models (LLMs). While existing evaluations primarily focus on whether a response aligns with a user's preferences, we argue that factuality is an equally important yet often overlooked dimension. In the context of personalization, we define a model as robust if its responses are both factually accurate and align with the user preferences. To assess this, we introduce PERG, a scalable framework for evaluating robustness in LLMs, along with a new dataset, PERGData. We evaluate fourteen models from five different model families using different prompting methods. Our findings show that current LLMs struggle with robust personalization: even the strongest models (GPT-4.1, LLaMA3-70B) fail to maintain correctness in 5% of previously successful cases without personalization, while smaller models (e.g., 7B-scale) can fail more than 20% of the time. Further analysis reveals that robustness is significantly affected by the nature of the query and the type of user preference. To mitigate these failures, we propose Pref-Aligner, a two-stage approach that improves robustness by an average of 25% across models. Our work highlights critical gaps in current evaluation practices and introduces tools and metrics to support more reliable, user-aligned LLM deployments.