The value alignment of large language models (LLMs) is crucial for ensuring responses align with human intention and value preferences. However, most evaluations of value alignment focus on Western or universal values, while assessments grounded in the value systems of specific countries remain scarce. In this paper, w...
We present Loopie, the most powerful open-source looped Transformer to date. The Loopie series consists of two Mixture-of-Experts (MoE) models: a 20B-parameter model with 2B active parameters and a 6B parameter model with 0.6B active parameters. Looped Transformers have long faced a challenge: given an N-fold pre-train...
Zitian Gao, Yilong Chen, Yihao Xiao et al.· 0 citations
Automatic prosodic segmentation identifies boundaries between speech units based on acoustic and linguistic evidence. Although recent deep learning approaches have achieved strong results for English, automatic segmentation for Brazilian Portuguese (BP) still relies largely on rule-based or traditional machine-learning...
R. Lima, J. Galdino, Marcos Vinícius Treviso· arXiv.org· 0 citations
Fine-tuned encoders deployed across heterogeneous NLP tasks face three compounding problems: mismatched inductive biases, class-imbalance corruption of feature statistics, and no mechanism to condition attention on external lexical knowledge. We introduce \textbf{\surgellm}, a unified transformer framework that address...
Noor Islam S. Mohammad, Ulug Bayazit· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Language-specific tokenizers improve tokenization quality and the downstream performance of models on those languages. Using one with an existing pretrained model requires either training a new model from scratch or adapting the vocabulary of the pretrained one. We propose Language-adaptive Maximum a Posteriori (LangMA...
Clara Meister, Suchir Salhan, Andrzej Szablewski et al.· 0 citations
Memory systems are essential for personalized Large Language Models (LLMs). However, retrieval within these systems primarily relies on semantic similarity, potentially missing logically critical memories with limited semantic overlap. We refer to this failure as semantic deflection, which remains insufficiently evalua...
Hongxun Ding, Xiang-Yu Yu, Cheng-Bing Wang et al.· 1 citation
Team science holds that leadership is contingent: it helps only under specific conditions. We ask the same question of multi-agent LLM teams: under what measurable conditions does process-level coordination control add value, and do those conditions match what team science predicts? This takes measurement, not just acc...
An answer from a vision-language model (VLM) to a medical image question can be accepted without review only under a threshold bounding the error of accepted answers, and certifying that bound costs local labels. We ask what such labels can certify for nine confidence estimators on five general-purpose open-weight VLMs...
Reza Khanmohammadi, Kundan Thind, Mohammad M. Ghassemi· 0 citations
Evaluating two LRMs against controls lacking the proposed recurrence or curriculum, it is found these patterns also appear in the controls and do not consistently affect behavior, and latent thoughts should be treated as hidden computation, not hidden explanation.
Darpan Aswal, Thomas Palmeira Ferraz, Yong-Xin Zhou et al.· arXiv.org· 7 citations· ⚡1
Large language models (LLMs) have shown impressive performance on diverse reasoning tasks, yet their capacity for structural reasoning in graphs remains unclear. We investigate whether LLMs can genuinely understand graph isomorphism -a fundamental problem in graph theory. While LLMs achieve near-perfect accuracy on iso...
PriFT (Prior-support guided Fine-Tuning), which derives token weights from a frozen pretrained reference to obtain a stable reweighting signal unaffected by fine-tuning, achieves state-of-the-art results among SFT baselines, while also providing a stronger initialization for subsequent RL.
Ke Wang, Shuang-Qi Li, Mathieu Salzmann et al.· 2 citations· ⚡2
Large language models (LLMs) can serve as helpful assistants, yet they can equally function as harm amplifiers that enable malicious users to achieve harmful outcomes beyond their capabilities through extended interactions. This risk manifests along two axes, i.e., democratizing domain expertise that allows novices to...
Flint-interactive gives AI agents a language for creating interactive visualizations. Click a chart, filter data, highlight related content, and explore beyond the view the author originally chose. The post Flint-interactive: A language for creating interactive data visualizations appeared first on Microsoft Research.
