搜索结果：LLMs

We discuss the results of using large language models (LLMs) to conduct original scientific research in an unfamiliar subject area during the Fall 2025 semester. Students in a graduate astronomy and astrophysics course were asked to test whether LLMs could help them complete research tasks faster and at a level of detail and accuracy required for scientific publication. Most students employed LLMs for a total of 5-10 hours. While all students completed a draft paper on an unsolved problem related to galaxies by semester's end, their impressions of the models' value varied. About half thought that the models saved them time. Many noted that LLMs failed to provide appropriately detailed insights or steps to addressing open, niche questions over a several-month timeframe. The LLMs also frequently (about 20% of the time) returned false citations, links, or summaries of papers. The models struggled with generating complex functional code, accessing online packages or Application Programming Interfaces (APIs), and retrieving astronomical datasets from existing archives. In writing code and in chats, the LLMs made implicit, overly simplifying assumptions and often doubled down even after

Large Language Models (LLMs) have become increasingly prominent for daily tasks, from improving sound-totext translation to generating additional frames for the latest video games. With the help of LLM inference frameworks, such as llama.cpp, which support optimizations such as KV-caching and quantization, it is now easier than ever to deploy LLMs on edge devices. Quantization is fundamental to enable LLMs on resource-constrained edge devices, and llama.cpp utilizes block floating point (BFP) quantization to drastically reduce the bit width of weights and input tensors, the memory footprint, and the computational power required to run LLMs. LLMs are typically quantized with mixed BFP quantization across the model layers to reduce the loss of model accuracy due to quantization. Therefore, to efficiently accelerate across the layers of BFP-quantized LLMs, specialized accelerators need to support different BFP variants without reconfiguration. To address this issue, we propose a Flexible Block FloatingPoint Quantization (F-BFQ) accelerator, which can dynamically switch between two BFP quantization variants and perform matrix multiplication (MatMul) operations. Our initial F-BFQ accele

We study syllogistic reasoning in LLMs from the logical and natural language perspectives. In process, we explore fundamental reasoning capabilities of the LLMs and the direction this research is moving forward. To aid in our studies, we use 14 large language models and investigate their syllogistic reasoning capabilities in terms of symbolic inferences as well as natural language understanding. Even though this reasoning mechanism is not a uniform emergent property across LLMs, the perfect symbolic performances in certain models make us wonder whether LLMs are becoming more and more formal reasoning mechanisms, rather than making explicit the nuances of human reasoning.

Social media platforms have become primary channels for health information in the Global South. Using gomutra (cow urine) discourse on YouTube in India as a case study, we present a post-facto Large Language Model (LLM)-assisted discourse analysis of 30 multilingual transcripts showing that promotional content blends sacred traditional language with pseudo-scientific claims in ways that sophisticated debunking content itself mirrors, creating a rhetorical register that LLMs, trained predominantly on Western corpora, are systematically ill-equipped to analyse. Varying prompt tone across three LLMs (GPT-4o, Gemini 2.5 Pro, DeepSeek-V3.1), we find that culturally embedded health misinformation does not look like ordinary misinformation, and this cultural obfuscation extends to gendered rhetoric and prompt design, compounding analytical unreliability. Our findings argue that cultural competency in LLM-assisted discourse analysis cannot be retrofitted through prompt engineering alone.

Large Language Models (LLMs) are nowadays prompted for a wide variety of tasks. In this article, we investigate their ability in reciting and generating graphs. We first study the ability of LLMs to regurgitate well known graphs from the literature (e.g. Karate club or the graph atlas)4. Secondly, we question the generative capabilities of LLMs by asking for Erdos-Renyi random graphs. As opposed to the possibility that they could memorize some Erdos-Renyi graphs included in their scraped training set, this second investigation aims at studying a possible emergent property of LLMs. For both tasks, we propose a metric to assess their errors with the lens of hallucination (i.e. incorrect information returned as facts). We most notably find that the amplitude of graph hallucinations can characterize the superiority of some LLMs. Indeed, for the recitation task, we observe that graph hallucinations correlate with the Hallucination Leaderboard, a hallucination rank that leverages 10, 000 times more prompts to obtain its ranking. For the generation task, we find surprisingly good and reproducible results in most of LLMs. We believe this to constitute a starting point for more in-depth stu

This is a summary of the paper "A Confederacy of Models: a Comprehensive Evaluation of LLMs on Creative Writing", which was published in Findings of EMNLP 2023. We evaluate a range of recent state-of-the-art, instruction-tuned large language models (LLMs) on an English creative writing task, and compare them to human writers. For this purpose, we use a specifically-tailored prompt (based on an epic combat between Ignatius J. Reilly, main character of John Kennedy Toole's "A Confederacy of Dunces", and a pterodactyl) to minimize the risk of training data leakage and force the models to be creative rather than reusing existing stories. The same prompt is presented to LLMs and human writers, and evaluation is performed by humans using a detailed rubric including various aspects like fluency, style, originality or humor. Results show that some state-of-the-art commercial LLMs match or slightly outperform our human writers in most of the evaluated dimensions. Open-source LLMs lag behind. Humans keep a close lead in originality, and only the top three LLMs can handle humor at human-like levels.

Recently, large language models (LLMs) have been deployed to tackle various software engineering (SE) tasks like code generation, significantly advancing the automation of SE tasks. However, assessing the quality of these LLM-generated code and text remains challenging. The commonly used Pass@k metric necessitates extensive unit tests and configured environments, demands a high labor cost, and is not suitable for evaluating LLM-generated text. Conventional metrics like BLEU, which measure only lexical rather than semantic similarity, have also come under scrutiny. In response, a new trend has emerged to employ LLMs for automated evaluation, known as LLM-as-a-judge. These LLM-as-a-judge methods are claimed to better mimic human assessment than conventional metrics without relying on high-quality reference answers. Nevertheless, their exact human alignment in SE tasks remains unexplored. In this paper, we empirically explore LLM-as-a-judge methods for evaluating SE tasks, focusing on their alignment with human judgments. We select seven LLM-as-a-judge methods that utilize general-purpose LLMs, alongside two LLMs specifically fine-tuned for evaluation. After generating and manually sc

Large language models (LLMs) are increasingly used for automated news credibility assessment, yet it remains unclear whether they apply even-handed standards across journalistic genres. We examine whether zero-shot LLMs are more likely to misclassify legitimate entertainment news as fake than legitimate hard news, using a within-dataset design on GossipCop from FakeNewsNet. Across four frontier models, we find a clear but model-specific genre asymmetry: DeepSeek-V3.2 and GPT-5.2 show false-positive-rate gaps of 10.1 and 8.8 percentage points, respectively (both $p < .001$), whereas Claude Opus 4.6 and Gemini 3 Flash show no comparable difference. A style-swap experiment yields only limited and inconsistent changes, suggesting that the asymmetry is not reducible to stylistic register alone. Prompt-based mitigation is likewise possible but not generic: framing the model as an entertainment-news fact-checker reduces false positives for DeepSeek-V3.2 by about 50\% without detectable recall loss, but offers little improvement for GPT-5.2. Exploratory qualitative coding further suggests two recurring error patterns in sampled false positives: treating private-life claims as inherently

Large Language Models (LLMs) are increasingly applied to tasks involving structured inputs such as graphs. Abstract Meaning Representations (AMRs), which encode rich semantics as directed graphs, offer a rigorous testbed for evaluating LLMs on text generation from such structures. Yet, current methods often arbitrarily linearize AMRs, discarding key structural cues, or rely on architectures incompatible with standard LLMs. We introduce SAFT, a structure-aware fine-tuning approach that injects graph topology into pretrained LLMs without architectural changes. We compute direction-sensitive positional encodings from the magnetic Laplacian of transformed AMRs and project them into the embedding space of the LLM. While possibly applicable to any graph-structured inputs, we focus on AMR-to-text generation as a representative and challenging benchmark. SAFT sets a new state-of-the-art on AMR 3.0 with a 3.5 BLEU improvement over baselines. Gains scale with graph complexity, highlighting the value of structure-aware representations in enhancing LLM performance. SAFT offers a general and effective pathway for bridging structured data and language models.

Large Language Models (LLMs) voice assistants are commonly built as cascaded Automatic Speech recognition (ASR) to LLM systems, where recognition errors can distort user intent. Dislikes may also arise from ambiguous, out-of-domain, or non-request turns, making it hard to isolate ASR effects. We release WASIL (it denotes connection or linking in Arabic): in-the-wild Arabic spoken interaction prompts with audio, ASR hypotheses, assistant responses, and explicit like/dislike feedback (8,529 turns; 14.2% dislikes), plus a 2,000-turn test set covering Modern Standard Arabic (MSA) and four major dialects with their labels. We provide low-cost gold transcripts via multi-ASR agreement-guided post-editing and annotate answerability (answerable, ambiguous/needs-clarification, unsupported, not-a-request/noise) to separate intrinsic unanswerability from ASR-induced degradation. Finally, we describe scalable reference-free evaluation of responses from ASR vs. gold transcripts using multi-judge LLM scoring.

In this paper, we report our experience with several LLMs for their ability to understand a process model in an interactive, conversational style, find syntactical and logical errors in it, and reason with it in depth through a natural language (NL) interface. Our findings show that a vanilla, untrained LLM like ChatGPT (model o3) in a zero-shot setting is effective in understanding BPMN process models from images and answering queries about them intelligently at syntactic, logic, and semantic levels of depth. Further, different LLMs vary in performance in terms of their accuracy and effectiveness. Nevertheless, our empirical analysis shows that LLMs can play a valuable role as assistants for business process designers and users. We also study the LLM's "thought process" and ability to perform deeper reasoning in the context of process analysis and optimization. We find that the LLMs seem to exhibit anthropomorphic properties.

Recent studies have applied large language models (LLMs) to machine translation quality estimation (MTQE) by prompting models to assign numeric scores. Nonetheless, these direct scoring methods tend to show low segment-level correlation with human judgments. In this paper, we propose a generation-based evaluation paradigm that leverages decoder-only LLMs to produce high-quality references, followed by semantic similarity scoring using sentence embeddings. We conduct the most extensive evaluation to date in MTQE, covering 8 LLMs and 8 language pairs. Empirical results show that our method outperforms both intra-LLM direct scoring baselines and external non-LLM reference-free metrics from MTME. These findings demonstrate the strength of generation-based evaluation and support a shift toward hybrid approaches that combine fluent generation with accurate semantic assessment.

The use of Large Language Models (LLMs) for generating Automated Planning (AP) models has been widely explored; however, their application to Hierarchical Planning (HP) is still far from reaching the level of sophistication observed in non-hierarchical architectures. In this work, we try to address this gap. We present two main contributions. First, we propose L2HP, an extension of L2P (a library to LLM-driven PDDL models generation) that support HP model generation and follows a design philosophy of generality and extensibility. Second, we apply our framework to perform experiments where we compare the modeling capabilities of LLMs for AP and HP. On the PlanBench dataset, results show that parsing success is limited but comparable in both settings (around 36\%), while syntactic validity is substantially lower in the hierarchical case (1\% vs. 20\% of instances). These findings underscore the unique challenges HP presents for LLMs, highlighting the need for further research to improve the quality of generated HP models.

Recent advances in Large Language Models (LLMs) are fostering their integration into several reasoning-related fields, including Automated Planning (AP). However, their integration into Hierarchical Planning (HP), a subfield of AP that leverages hierarchical knowledge to enhance planning performance, remains largely unexplored. In this preliminary work, we propose a roadmap to address this gap and harness the potential of LLMs for HP. To this end, we present a taxonomy of integration methods, exploring how LLMs can be utilized within the HP life cycle. Additionally, we provide a benchmark with a standardized dataset for evaluating the performance of future LLM-based HP approaches, and present initial results for a state-of-the-art HP planner and LLM planner. As expected, the latter exhibits limited performance (3\% correct plans, and none with a correct hierarchical decomposition) but serves as a valuable baseline for future approaches.

Humans are believed to perceive numbers on a logarithmic mental number line, where smaller values are represented with greater resolution than larger ones. This cognitive bias, supported by neuroscience and behavioral studies, suggests that numerical magnitudes are processed in a sublinear fashion rather than on a uniform linear scale. Inspired by this hypothesis, we investigate whether large language models (LLMs) exhibit a similar logarithmic-like structure in their internal numerical representations. By analyzing how numerical values are encoded across different layers of LLMs, we apply dimensionality reduction techniques such as PCA and PLS followed by geometric regression to uncover latent structures in the learned embeddings. Our findings reveal that the model's numerical representations exhibit sublinear spacing, with distances between values aligning with a logarithmic scale. This suggests that LLMs, much like humans, may encode numbers in a compressed, non-uniform manner.

Misinformation such as fake news and rumors is a serious threat on information ecosystems and public trust. The emergence of Large Language Models (LLMs) has great potential to reshape the landscape of combating misinformation. Generally, LLMs can be a double-edged sword in the fight. On the one hand, LLMs bring promising opportunities for combating misinformation due to their profound world knowledge and strong reasoning abilities. Thus, one emergent question is: how to utilize LLMs to combat misinformation? On the other hand, the critical challenge is that LLMs can be easily leveraged to generate deceptive misinformation at scale. Then, another important question is: how to combat LLM-generated misinformation? In this paper, we first systematically review the history of combating misinformation before the advent of LLMs. Then we illustrate the current efforts and present an outlook for these two fundamental questions respectively. The goal of this survey paper is to facilitate the progress of utilizing LLMs for fighting misinformation and call for interdisciplinary efforts from different stakeholders for combating LLM-generated misinformation.

Large Language Models (LLMs) have demonstrated remarkable performance across various disciplines and tasks. However, benchmarking their capabilities with multilingual spoken queries remains largely unexplored. In this study, we introduce SpokenNativQA, the first multilingual and culturally aligned spoken question-answering (SQA) dataset designed to evaluate LLMs in real-world conversational settings. The dataset comprises approximately 33,000 naturally spoken questions and answers in multiple languages, including low-resource and dialect-rich languages, providing a robust benchmark for assessing LLM performance in speech-based interactions. SpokenNativQA addresses the limitations of text-based QA datasets by incorporating speech variability, accents, and linguistic diversity. We benchmark different ASR systems and LLMs for SQA and present our findings. We released the data at (https://huggingface.co/datasets/QCRI/SpokenNativQA) and the experimental scripts at (https://llmebench.qcri.org/) for the research community.

In this work, we argue that large language models (LLMs), though trained to predict only the next token, exhibit emergent planning behaviors: $\textbf{their hidden representations encode future outputs beyond the next token}$. Through simple probing, we demonstrate that LLM prompt representations encode global attributes of their entire responses, including $\textit{structure attributes}$ (e.g., response length, reasoning steps), $\textit{content attributes}$ (e.g., character choices in storywriting, multiple-choice answers at the end of response), and $\textit{behavior attributes}$ (e.g., answer confidence, factual consistency). In addition to identifying response planning, we explore how it scales with model size across tasks and how it evolves during generation. The findings that LLMs plan ahead for the future in their hidden representations suggest potential applications for improving transparency and generation control.

Large Language Models (LLMs) have demonstrated remarkable text generation capabilities, and recent advances in training paradigms have led to breakthroughs in their reasoning performance. In this work, we investigate how the reasoning effort of such models scales with problem complexity. We use the infinitely scalable Tents puzzle, which has a known linear-time solution, to analyze this scaling behavior. Our results show that reasoning effort scales with problem size, but only up to a critical problem complexity. Beyond this threshold, the reasoning effort does not continue to increase, and may even decrease. This observation highlights a critical limitation in the logical coherence of current LLMs as problem complexity increases, and underscores the need for strategies to improve reasoning scalability. Furthermore, our results reveal significant performance differences between current state-of-the-art reasoning models when faced with increasingly complex logical puzzles.