(c) 2026 Pranay Mahendrakar. Licensed under CC BY 4.0. The expectation that language models will keep improving after the supply of human text runs out rests on a claim about synthetic data: that a model, or a family of models, can produce training text worth more than the text it was trained on. A competing reading ho...
Pranay M. Mahendrakar· Zenodo (CERN European Organi...· 0 citations
In a typical systematic paper, specimen morphology is described with natural language, and taxonomic specialists use glossaries of technical terminology to establish and communicate taxon concepts, using morphology to delimit species. Textual descriptions of specimen morphology are therefore foundational to biological...
Luke Mander, Peng Sun, S. Punyasena et al.· Biology Letters· 0 citations
Grounded spatial reasoning about indoor scenes rests on a geometric reconstruction that the multimodal system does not itself produce. Feed-forward multi-view reconstruction models operate under a fixed view budget, making frame selection a critical upstream decision. We show that under this setting the conventional pr...
Science education requires learning resources that connect scientific knowledge with contextual problem solving, environmental responsibility, and learners’ cultural experiences. This study aimed to determine the validity and practicality of the Pancaniti ethnoscience-based UA-EVAA (Understanding, Awareness, Evidence-B...
Tourism firms at heritage destinations must manage operational risks and protect the resources on which their businesses depend. Online reviews could inform managerial follow-up, but sparse mentions and high ratings may limit their value. This study examines these limits through a transparent business analytics pipelin...
Q. Le, Hien Minh Nguyen, Kiet Tuan Dinh· European Journal of Manageme...· 0 citations
The widespread adoption of large language models has increased the need to distinguish human-authored text from AI-generated content, including cases in which generated text is subsequently rewritten or paraphrased. In practice, many detectors output only a single AI-likelihood score or percentage, offering limited tra...
Mining environmental metagenomics data, the authors show that small HK97-fold-encoding viral genomes bridge Duplodnaviria viral capsids and cellular protein shells, revealing bidirectional evolution in which viral procapsids enable the transitions between viral and metabolic compartments.
Abelardo Aguilar, Ingrid Miranda-Pérez, Omer Nadel et al.· Nature Communications· 0 citations
Rising DAS adoption was accompanied by a more heterogeneous pooled composition of sharing mechanisms rather than a single pathway towards open data, a shift driven largely by the staggered entry of journals.
Hiroyuki Tsunoda, Yuan Sun, Masaki Nishizawa et al.· Proceedings of the 26th ACM/...· 0 citations
This dissertation asks whether compression can be made capability-aware by borrowing the unit of analysis from mechanistic interpretability, and presents the problem formulation, the gap identified in a targeted review of the compression and interpretability literatures, and the design of a first experiment measuring c...
Farshad Ghorbanishovaneh· Proceedings of the 26th ACM/...· 0 citations
A preliminary task-level comparison of a Direct Generation Workflow and a Decomposed Generation Workflow under matched inputs and evaluation conditions provides preliminary evidence that the value of multi-agent decomposition is task-dependent.
A. Hossain, M. Chowdhury, Shafayet Nur et al.· Proceedings of the 26th ACM/...· 0 citations
The loss of fidelity shows that intermediate teacher evaluations provide useful supervision for approximating the evaluative function, and whether reducing the influence of intermediate teacher evaluations improves or impairs approximation is tested by comparing plain binary cross-entropy with weighting, masking, and h...
William A. Ingram, Edward A. Fox· Proceedings of the 26th ACM/...· 0 citations
Long-lived technical organizations depend on senior subject matter experts whose knowledge is only partially preserved in reports, publications, and archival metadata. Digital libraries preserve those records, but they rarely provide centralized, expert-scoped, question-answering access to the knowledge of people who l...
James Ecker, Jian Wu, B. D. Allen et al.· Proceedings of the 26th ACM/...· 0 citations
Assistant Professor Benjamin Lindquist studies how engineers deliberately injected randomness into early AI, giving rise to deep-learning models that sometimes hallucinate.
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.
