Looped language models increase effective depth by repeatedly applying a shared block of layers, but existing large-scale recipes require multi-stage training over trillions of tokens, while the benefits of recurrence remain difficult to separate from differences in data and training. In this work, we establish practic...
Andrei Marchenko, В. Е. Безруков, Oleg Kashurin et al.· 3 citations· ⚡1
The performance of AI coding agents is highly dependent on their underlying coding rules. However, existing coding rules are typically hand-crafted and fixed, making the process labor-intensive and often suboptimal. In this work, we propose RuleEvolve, a self-evolving framework for coding rules. RuleEvolve maintains a...
Zheng-Yuan Jiang, Reachal Wang, Yue-Peng Hu et al.· 0 citations
Large Language Models increasingly serve as interfaces for knowledge-intensive information seeking tasks across languages by synthesizing multilingual evidence. Prior work has shown that they often exhibit query-language preference -- the tendency to favor sources written in the language of the query -- but has largely...
Dayeon Ki, Ruochen Zhang, Silviu Cucerzan et al.· 0 citations
Large language models are characterized by three key properties: capability, alignment, and faithfulness. Prior work studies the tradeoffs between capability and alignment, and between capability and faithfulness, but a third tension remains underexplored: the alignment-faithfulness conflict. We show that aligned model...
Pardis Sadat Zahraei, Janvijay Singh, Gokhan Tur et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Longitudinal clinical reasoning requires large language models (LLMs) to identify and integrate relevant evidence distributed across extended patient histories. Although long-context models can process increasingly large amounts of information, providing more history does not necessarily make relevant evidence more acc...
Taye Akinrele, Noorbakhsh Amiri Golilarz, Subash Neupane et al.· 0 citations
In-context learning (ICL) can be amortized into latent objects (task vectors, function vectors, context vectors) that recover few-shot behavior at zero-shot inference cost, but recent theory shows a static vector acts as a single synthetic demonstration and must fail on high-rank mappings such as word-level bijections....
EurekaBench is introduced, a cross-domain benchmark that tests AI agents'ability to conduct long-horizon experiments and discover mechanisms that explain observations, and shows that current AI agents often overly fixate on predictive accuracy optimization, surpassing human scientists, while falling substantially short...
Jia-Yi Geng, Zhengxuan Wu, Kevin S. Chen et al.· 1 citation
Replay selectors often rank cached trajectories by format feedback, confidence, freshness, or response length, although cache-level correctness and downstream learner utility are distinct objectives. We formalize this selection-to-learning gap and introduce FAER as an auditable full-trajectory replay framework. Its tra...
Miaobo Hu, Shuhao Hu, Xiaobo Guo et al.· 0 citations
Analytical charts in multimodal deep research encode quantitative claims, requiring every visualized value to be faithfully grounded in supporting evidence. Unlike retrieved images that mainly provide contextual information, charts require numerical fidelity: visualized values should not only match retrieved evidence q...
Yuxin Yue, Yingchen Zhang, Ruqing Zhang et al.· 0 citations
Laurito et al. (PNAS 2025) showed that large language models choosing between two descriptions of the same product, paper or film prefer the description written by a language model over the one written by a person, by a wide margin over what human judges do. Their design crosses five generators with the same five model...
Multimodal on-policy distillation (OPD) aims to improve visual grounding while preserving the strong reasoning capabilities of language models. Recent multi-teacher approaches combine LLM and VLM teachers to provide complementary supervision. However, directly using a VLM's full predictive distribution entangles its vi...
Jaeyun Shin, Hangeol Chang, Jong Chul Ye· 0 citations
Tool-using agents can expose citations and execution logs while leaving a critical association unaudited: whether the claim shown to a user is the claim emitted by the committed execution and supported by the cited source. A valid citation and a valid trace can therefore remain individually well formed while being tran...
Miao-Bo Hu, Shu-Hao Hu, Xiao-Bo Guo et al.· 0 citations
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.
