PAPER PLAINE

Permission systems are the backbone of security in cloud platforms and distributed software. If revoking access doesn't actually stop all paths to action—like a merge happening with the wrong code, or a message publishing after removal—attackers can exploit the gap between when you think something is locked down and when it actually is. The fixes tested on Kafka show that proper closure checking can block these risks without slowing down legitimate operations.

AI-driven trading and financial services are growing, but if an AI agent can both decide and execute a trade, a single bug or adversarial prompt could drain an account or lock up client money before anyone notices. This system creates an enforced separation: the AI proposes, but humans and pre-set rules decide whether the proposal actually runs. Banks and fintech firms adopting this approach get an auditable record of *why* each trade was approved or blocked, reducing rogue-agent risk and making it possible to prove later who was responsible for any losses.

Training large language models is expensive, and knowing how to divide limited annotation resources between SFT and RL can save significant cost without sacrificing performance. This approach lets teams run quick experiments on small models to confidently set their budget split for large models, avoiding wasteful full-scale trial-and-error. The method also adapts when the cost difference between SFT and RL data changes, making it practical for real-world budget constraints.

Many real control systems face uncertainty that changes with their own state or decisions — a robot's collision risk depends on where it is, a satellite's tumbling depends on its spin rate. Previous methods either couldn't handle this, required custom solvers that didn't scale up, or only worked for simple cases. This approach uses standard software and applies broadly, making it practical to design safer, more robust controllers for aerospace, robotics, and other safety-critical applications.

Analog circuit design currently requires extensive trial-and-error and knowledge locked behind non-disclosure agreements, slowing innovation and limiting who can participate. This database removes those barriers by making proven designs reusable and machine-readable, enabling AI tools to find and adapt existing solutions automatically. For companies and researchers, that means faster design cycles and fewer errors; for the field, it means knowledge can accumulate and compound rather than staying trapped in individual labs.

As companies deploy larger language models on phones, servers, and edge devices, quantization compression is essential to cut costs and latency. This finding changes how engineers should allocate limited precision budgets: instead of hunting for problem layers, spreading improvements evenly across the model works better in practice and is simpler to implement. This could reduce the engineering overhead of tuning quantization for each new model.

Machine learning teams routinely use synthetic data to fix imbalanced datasets, but have lacked clear guidance on which generation methods work best. This framework provides a principled way to evaluate synthetic data quality before deploying it, potentially saving teams from investing time in high-fidelity generators that won't improve actual performance. It also shows that simpler oversampling methods are sometimes more reliable than complex generative models—a finding that could reshape how practitioners choose their augmentation strategy.

Drug discovery relies on predicting how experimental molecules will bind their targets before expensive testing begins. This work exposes a critical gap: most peptide-binding models are benchmarked in unrealistic conditions and would fail in the real world, where researchers must predict binding to new disease targets they've never seen before. The findings demand that future benchmarks test what actually matters—whether methods can handle unfamiliar targets—and reveal which representation strategies hold up under that harder test.

Language models deployed to millions of users encounter rare failure modes that standard testing never catches. BLOOM-WILT makes it cheap and practical to continuously hunt for these hidden safety problems after deployment, meaning developers can catch and fix harms that would otherwise slip through to real users. The system's rankings also revealed that some models previously thought safer than others actually weren't—a correction that affects which systems get deployed.

Heavy-ion collisions at facilities like the Large Hadron Collider create extreme conditions that reveal fundamental physics, but are hard to simulate because particles don't forget their past interactions instantly. This framework opens the door to using next-generation quantum computers to model these collisions more accurately, potentially unlocking new insights into how matter behaves under the most violent conditions in nature.

As companies increasingly hand off coding work to AI agents pulling from repositories they don't fully control, understanding how simple invocation choices affect safety is critical. A developer asking for test execution instead of general setup could expose their system to attack 4.5 times more, and they'd likely never notice. This work gives teams concrete guidance on safer ways to phrase their requests and which tasks pose the highest risk.

Language models spend enormous compute power converting every intermediate thought into discrete tokens, even though these internal reasoning steps are never seen by users. This work shows those internal steps don't need to be tokenized at all, potentially making AI systems faster and cheaper to run while improving their ability to solve hard problems.