Multimodal Machine Translation aims to incorporate additional signal from non-textual modalities to improve translations by resolving ambiguities. While models, through multimodal fusion, are able to accept images related to the source text, they can ignore this information. Therefore, increasing their visual sensitivi...
Pawe{\l} M\k{a}ka, Piotr Andruszkiewicz, Yusuf Can Semerci et al.· 0 citations
JevAdvBench is introduced, to the authors' knowledge the first adversarial benchmark for RLCD models, with 812 typed questions over 66 scenarios, and a black-box attack suite of 9,744 single-edit variants that each edit one part of a request, with billed input tokens confirming that the edit reached the model.
Jian-Yi Hu, Hang Zhang, Yi Liu et al.· 12 citations· ⚡3
G$^2$PTQ is presented, a unified PTQ framework with Generalized Gradient Compensation that integrates both first- and second-order information under a globally supervised, block-wise optimization objective and enables better alignment with the full-precision model, outperforming state-of-the-art baselines.
Rui-Kang Liu, Hao-Li Bai, Yuxuan Sun et al.· 0 citations
It is shown that the population-optimal UDM predictor generally depends on time: time controls how much the model should trust the observed context, and this dependence can become negligible in finite-data settings relevant to language.
Chunsan Hong, Chieh-Hsin Lai, Satoshi Hayakawa et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
EoupCT effectively preserves both task-specific proficiency and inherent general-purpose knowledge, successfully mitigating the catastrophic forgetting, and forms a multi-objective optimization problem and introduces a first-order efficient Pareto optimizer.
Bing Wang, C. Li, Xin-Qiang Cai et al.· 1 citation
Persistent-negative adversarial distillation consistently improves performance over current methods at matched discriminator compute and yields smoother fresh-policy discriminator trajectories, with fewer below-chance dips.
Hai-Xu Ma, S. Lahrichi, Wei-Wei Li et al.· 0 citations
When a language model receives two conflicting documents as input, how does it decide which one to prioritize? Does it rely on how the sources are framed or the presentation order of the documents? We evaluated this behavior on Google's pre-trained Gemma 4-e4b model across a targeted behavioral suite (n = 13 items, 784...
Existing computer-use agent benchmarks do not fully evaluate agents acting as assistants. A useful assistant retrieves information across complex, multi-step workflows, synthesizes it into artifacts (documents, presentations, spreadsheets), and navigates program interfaces to produce a coherent final product. Such work...
Alexander Gill, Md Farhan Ishmam, X. Nguyen et al.· 1 citation
MemProbe, a cognitive-science-inspired framework for diagnosing stability-plasticity tradeoffs in agent memory, is introduced and it is shown that systems with similar aggregate scores exhibit distinct behavioral profiles.
Inquesto Score (IS), a protocol for measuring voice-agent reliability as the percentage of calls in a fixed, versioned evaluation population that achieve the caller's goal without a functional failure or worse, is introduced.
These results establish persona-set geometry as a task-agnostic mechanism for eliciting divergent LLM outputs, and support persona diversification as a reusable complement to prompt optimization.
Sang-Bin Moon, Nicole Cho, Daniel Borrajo et al.· 0 citations
This work proposes CARGO, a framework that treats retrieved references as procedural exemplars and grounds factual judgments in the live instance's observed context and gates evaluation by retrieval confidence, casting production evaluation as selective prediction.
Mukul Chhabra, Shail Patel, Luigi Medrano· 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.
A new computational approach reveals how cells communicate to assemble into tissues. This could accelerate stem cell engineering, improve disease models, and inform therapies.
Microsoft Research Blog· microsoft.comOct 15, 2026
Zero-knowledge proofs can verify computing results without rerunning workloads or exposing sensitive data. Magnetar helps bring them into the mainstream with support for standard programming languages, major speed gains, and quantum-resistant security. The post Proof-based general-purpose computing with Magnetar appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduSep 24, 2026
