Tiny flaws in diamonds could unlock a quantum computing revolution
Scientists have discovered a faster way to identify diamond-based materials that could power quantum computers and sensors at room temperature. The breakthrough has already revealed promising new quantum building blocks that could help unlock computing capabilities far beyond those of today's superc...
SynthCorp: A Platform for Creating Benchmarks for Data Agents
arXiv:2611.14765v1 Announce Type: new
Abstract: Current benchmarks for LLM-driven data agents saturate quickly and cannot target specific capabilities. We present SynthCorp, a platform that creates customizable data estates and benchmarks by simulating how business events produce and modify corpora...
When Should Semantic Priors Enter Partially-Identified Causal Discovery?
arXiv:2611.14736v1 Announce Type: new
Abstract: Observational causal discovery often leaves causal di- rections unresolved or unstable. Large language mod- els (LLMs) can provide semantic information to resolve these directions, but incorrect proposals can exclude the true causal graph and invalida...
Together Compiler: Continual Compilation for AI Inference
arXiv:2611.14681v1 Announce Type: new
Abstract: Today's AI workloads are rapidly changing: new models are released more often, model architectures are increasingly diverse, and traffic distributions change frequently. The challenge we face at Together AI is continual optimization for these changing...
Cleaning an Agent's History Hurts When the World Still Holds the Error
arXiv:2611.14654v1 Announce Type: new
Abstract: Tool-using language-model agents accumulate histories of their own mistakes, and a system that compacts, repairs or hands off such a history decides what it says about those mistakes while their effects stay in the world. Prior work cleans the history...
What's new in research at Microsoft? Host Anson Ho explores the latest, including AI-powered universal translation, a breakthrough in a 31-year-old mathematics challenge, and advances making AI coding agents faster and more efficient.
The post Microsoft Research Focus – Edition #98 appeared first o...
Inside this quantum computer, memory vibrates like a guitar string
Researchers have created a quantum computer that stores information in microscopic vibrations, potentially making quantum machines smaller, more powerful, and better at preserving information. The breakthrough successfully demonstrated advanced quantum calculations and could help overcome major barr...
Language model (LM) alignment improves model outputs to reflect human preferences while preserving the capabilities of the base model. The most common alignment approaches are (i) reinforcement learning, which maximizes the expected reward under a KL-divergence constraint, and (ii) best-of-NNN align...
Auditable Human-Agent Teams for Autonomous X-ray Experimentation
arXiv:2611.12933v1 Announce Type: new
Abstract: Closed-loop autonomous experimentation can accelerate scientific discovery, but implementation effort limits accessibility. Agentic systems may lower this barrier, although sustained operation on physical instruments remains largely unexplored. Here, ...
An ensemble of graph generative models for assisted history matching of the Brugge field case
arXiv:2611.12927v1 Announce Type: new
Abstract: Assisted history matching remains non-unique because many geological realisations can reproduce equivalent production responses with substantial differences between individual well matches. This ambiguity increases when a field comprises several depos...
Predictive Surprise as Self-Grounding Concept Bottleneck for Interpretable Time Series
arXiv:2611.12955v1 Announce Type: new
Abstract: Time-series classifiers deployed in risk-sensitive domains must be (1) accurate, (2) interpretable, and (3) able to abstain when evidence is weak. Existing methods do not deliver all 3 at once, and the obstacle is structural, rather than logistical. M...
Exploration via Exploitation: The Blessing of Reward Diversity in Personalized Federated RL
arXiv:2611.12964v1 Announce Type: new
Abstract: While federated reinforcement learning (RL) has traditionally guaranteed high efficiency in homogeneous environments, collaboration is often hindered when agents possess distinct goals. In this paper, we flip this paradigm, framing reward heterogeneit...
StatPrompt-LLM: A Lightweight Statistical Refinement for Foundation-Style Time-Series Forecasting
arXiv:2611.12974v1 Announce Type: new
Abstract: Time-series forecasting has recently seen a paradigm shift with the introduction of foundation-style models that adapt large pretrained language architectures to sequential data. However, effectively exposing fundamental temporal structures, such as t...
EarthRecord: From Pixels to Land Change with Multimodal Agents over Four Decades of Satellite Imagery
arXiv:2611.12934v1 Announce Type: new
Abstract: While multimodal agents now carry out complex data analysis with tools, their ability to turn decades of satellite imagery into reliable land-change analysis, crucial for Earth observation science, remains unmeasured. Existing benchmarks fall short: t...
Train on prefill, monitor at decode: Using sparse autoencoders for topic monitoring
arXiv:2611.12935v1 Announce Type: new
Abstract: Topic monitoring is the task of labelling the topic of each token in a model's response during generation. Activation-based monitors that run at generation time have so far been trained to detect harmful content and are limited to two or three classes...
Tracing Psychometric Inference in Large Language Models
arXiv:2611.12937v1 Announce Type: new
Abstract: Do large language models represent psychological constructs internally, or do they only generate outputs that resemble psychometric structure? We address this question using a cross-persona paradigm across 14 model variants from two families (Llama 3 ...
LandslideFM: A Foundation Model Towards Unified Landslide Perception and Understanding
arXiv:2611.12936v1 Announce Type: new
Abstract: Intelligent landslide detection is essential for disaster prevention. However, existing models primarily focus on isolated visual tasks, require extensive labeled data, and lack deep analytical and scene-understanding capabilities. To address these li...
Research Teams Mine the Intersection of Neuroscience and AI
Wu Tsai Neuro and HAI are again teaming up to support efforts to better treat psychiatric disease, understand brain development, and improve AI models with inspiration from human experience.
Lazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents
Microsoft researchers have discovered “Lazarus,” a backdoor technique that can reprogram dormant neurons in an AI model, causing an otherwise normal AI agent to execute malicious commands when a hidden trigger appears. The threat is especially serious for AI agents with tool or terminal access. A tr...
Your AI agent's brain is 72% decoration. Here's the audit
Theory of mind, causal reasoning, global workspace: a recent arXiv audit found most named modules in one LLM agent were decoration. Rewiring two of them took a planner to a perfect score.
META-IR: Meta-Introspective Reasoning for Adaptive Safety Alignment in LLM
arXiv:2611.10089v1 Announce Type: new
Abstract: Large Language Models (LLMs) remain vulnerable to jailbreak attacks and harmful-content prompts, particularly under adversarial distribution shifts where static alignment pipelines fail to generalize. Although introspective methods such as STAIR stren...