PAPER PLAINE

Trucking companies must decide in seconds whether to accept shipping jobs, weighing whether they can physically reach pickups, what future opportunities they'll miss, and how much repositioning will cost. Researchers built FreightBidBench, the first public testing ground for this problem, and showed that a smart decision-making system can match the performance of perfect hindsight 98% of the time while cutting decision time nearly in half.

Training image generation models to match human preferences currently wastes feedback by treating all learning moments equally. This paper shows that some timesteps in the generation process carry much more useful information than others, and some past examples are more worth revisiting — together, these insights reduce the amount of human feedback needed by up to sixfold while maintaining quality.

When multiple identical four-level atoms are arranged in certain geometric configurations, they emit light in coordinated bursts that grow much faster than expected—following a mathematical power law where the peak brightness scales roughly with the square of the number of atoms rather than linearly. Researchers developed a unified mathematical framework using group theory to predict and explain these synchronized emission patterns across seven different atomic geometries.

Researchers built a system where multiple AI agents work together on long drug-development projects while staying within memory limits by forgetting completed tasks. The system consistently picked the right mathematical models for predicting how drugs move through the body, even when switching between cheaper and more powerful AI models, and recovered drug parameters more accurately than single-agent approaches.

Vision-language models often make up false explanations about how physical interactions will unfold, and their reasoning doesn't match their actual behavior. Researchers developed VAORA, a reward system that forces AI to ground its reasoning in what it actually sees and does, significantly improving the model's ability to handle new tasks and unfamiliar environments.

Matching security regulations to actual technical controls in cloud systems has always meant manual work by compliance experts. This paper trains AI models on 14,000 examples of security requirements paired with technical metrics, allowing them to automate the matching process—with the best models improving accuracy by up to 23 percentage points over standard AI approaches. The technique also works across different security standards, so a mapping learned from one regulation can help with others.

Satellites in lunar orbit bunch together overhead at the south pole, creating poor positioning geometry—even a 12-satellite constellation barely matches Earth GPS performance. Adding just three ground beacons on high terrain around the pole fixes the problem almost completely, dropping positioning error tenfold and costing far less than launching extra satellites.

When AI systems answer complex questions by hopping between related facts in a knowledge graph, they often get stuck because intermediate steps use different words than the original question. RSF-GLLM solves this by first tracing a path through the graph using meaning-based relevance scores rather than word matching, then using that concrete path to guide a language model toward the right answer—achieving competitive accuracy while running significantly faster than similar systems.

TabPack trains many slightly different AI models in parallel on tabular data and automatically picks the best ones, all without requiring tedious manual tuning beforehand. The method performs as well as carefully hand-tuned competitors while using far fewer computing resources—in one test, it completed on a MacBook faster than a baseline method running on a high-end GPU.

A new method called TREK helps AI models solve difficult math and reasoning problems by first learning verified solutions from a stronger teacher, then refining its own approach. On challenging math competitions, the method improved Qwen3-8B's score on AIME 2025 from 36.9 to 40.3, and lifted performance on agent-based tasks like virtual world navigation from 75.8 to 82.8 percent success.

Researchers developed a control system that uses live weather prediction data to compute small atmospheric tweaks designed to reduce precipitation. By treating the weather forecast model as a constantly updating guide and solving for the most efficient adjustments at each time step, the system achieved significant rainfall reduction even when simpler methods failed — and did so with much less computer time than full-horizon planning approaches.

Researchers created a new watermarking system for AI-generated text that lets you hide information inside it—then reveal only parts of that hidden message to different people. Unlike older methods that force you to expose everything or nothing, this approach uses a hierarchical system that gives each reader access to only what they're authorized to see, without degrading the quality of the generated text.