KUPAS MASTER, an experience engineering platform built around nine-layer cognitive corpus construction, turns heterogeneous work records and practitioner interviews into traceable, reusable experience corpora for agents, and provides a practical path from individual tacit experience to organizational knowledge and agen...
Chang-Mian Wang, Yu-Chao Ma, Xu-Chao Lu et al.· 0 citations
This paper recast context compression as a causal decision preservation problem over discrete interaction units and introduces FOCUS, a training-free context compression framework that operates entirely at test time that requires no offline data collection or fine-tuning.
Shantanu Dixit, Anson Bastos, Xu-Chao Zhang et al.· 0 citations
Users of language-based assistive agents often make ambiguous requests. In response, an assistant can either directly act on its interpretation of the request --- risking misalignment with the user --- or ask a clarifying question. Which option is the most safe and helpful? A common approach is to ask questions that mi...
T. Nguyen-Hien, Y. Teh, Wee-Sun Lee et al.· 0 citations
On-policy distillation, where a student learns from a stronger teacher's feedback on its own outputs, is a common way to pass reasoning to smaller models. We analyze what it transfers at small scale, distilling Qwen3-8B into Qwen3 4B, 1.7B and 0.6B students, in thinking mode (reason at length, then end the reasoning an...
Hong-Yang Li, Yi-Ming Zhu, Xiao Li et al.· 1 citation
Reach audiences
Advertise in front of researchers, engineers, and readers.
An agent that uses tools typically responds to what the user explicitly asks, yet completing the task may require information the user never requested. Work on proactive agents mainly studies whether and when an agent should act on its own, not what information it should pursue. We study a distinct axis of proactivity:...
Ido Levy, Asaf Yehudai, Segev Shlomov et al.· 0 citations
CoEM preserves potentially useful source excerpts verbatim in a pending set, allowing subsequent context to clarify their relevance before irreversible compression, and consistently improves long-context reasoning.
Jing-Guang Li, Yebo Wu, Zu-Yi Guo et al.· 0 citations
This work introduces a preference-oriented benchmark and systematically characterize the capabilities and limitations of personal agents associated with these three dimensions of harness evolution, offering a unified perspective on the limits of personalization through harness evolution and inform future harness design...
This work proposes self-instructing policy optimization (SIPO) with a contrastive self-teacher to provide dense credit, and demonstrates that SIPO outperforms both RLVR and OPSD baselines without an external teacher or additional generation.
Zhenrui Yue, Hui-Min Zeng, Yue-Qi Wang et al.· 0 citations
LibraryDesignBench is introduced, a two-phase benchmark in which an agent implements a full-featured library from a specification that defines required capabilities and potential use cases without prescribing the design.
Gabriel Orlanski, Alex L. Zhang, Avi Trost et al.· 0 citations
Pretrained transformers use little of their depth to follow references in context. Thirteen base models reliably follow only 1.4-3.6 lines, and extra pretrained loops add little. A task-trained rank-8 LoRA at one early layer extends this computation with all model weights frozen. Qwen3-8B improves from 15.5% to 99% exa...
The Optimal Transport-based Robust Off-Policy Evaluation (OTROPE) is proposed, a likelihood-free evaluation that performs distributional correction in a semantic space via optimal transport to align labeled behavior-policy samples with unlabeled target-policy samples, yielding a doubly robust-style evaluation.
REST (REpresentation-Supervised Thoughts), a training objective that turns four properties of a valid thought representation into differentiable losses added to CE, instantiate in latent single-agent and multi-agent systems, without architectural changes or added parameters at inference.
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
A new method, called CW-Net, translates the reasoning process of an autonomous vehicle’s AI system into understandable concepts that explain its behavior.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
