With help from artificial intelligence, MIT researchers have designed novel antibiotics that can combat two hard-to-treat infections: drug-resistant Neisseria gonorrhoeae and multi-drug-resistant Staphylococcus aureus (MRSA). Using generative AI algorithms, the research team designed more than 36 million possible compounds and computationally screened them for antimicrobial properties. The top candidates they discovered are structurally distinct from any existing antibiotics, and they appear to work by novel mechanisms that disrupt bacterial cell membranes. This approach allowed the researchers to generate and evaluate theoretical compounds that have never been seen before -- a strategy that they now hope to apply to identify and design compounds with activity against other species of bacteria. "We're excited about the new possibilities that this project opens up for antibiotics development. Our work shows the power of AI from a drug design standpoint, and enables us to exploit much larger chemical spaces that were previously inaccessible," says James Collins, the Termeer Professor of Medical Engineering and Science in MIT's Institute for Medical Engineering and Science (IMES) and Department of Biological Engineering, and a member of the Broad Institute.

"These days at 70 years old you are still a child," China's Xi Jinping, 72, was translated as saying. "With the developments of biotechnology, human organs can be continuously transplanted, and people can live younger and younger, and even achieve immortality," Russia's Vladimir Putin, also 72, is reported to have replied. In reality, rounds of organ transplantation surgery aren't likely to help anyone radically extend their lifespan anytime soon. And it's a simplistic way to think about aging--a process so complicated that researchers can't agree on what causes it, why it occurs, or even how to define it, let alone "treat" it. This article first appeared in The Checkup, MIT Technology Review's weekly biotech newsletter.

The giant, remote-controlled vehicle -- somewhere between a tractor trailer, a tank and a Zamboni in appearance -- slowly rolled across the dry, brittle grass growing between the tangle of freeways making up the 101 and 23 interchange in Thousand Oaks. And as it rolled over the land, that fire incinerated any brush it encountered, leaving only a thin smoke cloud billowing from the top of the machine, some flashes of orange and red from behind its metal skirt and, in its wake, a desolate, smoldering black line. BurnBot isn't the fastest way to rid a landscape of dangerously flammable vegetation (it tops out at around 0.5 mph) but it can do something that traditional vegetation management techniques cannot: with almost surgical precision, it can kill the flammable brush sitting within feet of homes and highways on even the hottest and driest days and with virtually no safety risks or disruptions to daily life. On a recent summer afternoon, as wildland firefighters maneuvered the machine and mopped up the charred earth on a stretch of highway about 30 miles west of Los Angeles on the 101, a who's who of SoCal's wildfire leadership looked on -- from the California Department of Forestry and Fire Protection, local fire departments, Caltrans, the U.S. forest and park services, Southern California Edison and state Legislature. The sweet smoky smell of wildland fire permeated the hot midday air.

Differential privacy (DP) is the standard for privacy-preserving analysis, and introduces a fundamental trade-off between privacy guarantees and model performance. Selecting the optimal balance is a critical challenge that can be framed as a multi-objective optimization (MOO) problem where one first discovers the set of optimal trade-offs (the Pareto front) and then learns a decision-maker's preference over them. While a rich body of work on interactive MOO exists, the standard approach -- modeling the objective functions with generic surrogates and learning preferences from simple pairwise feedback -- is inefficient for DP because it fails to leverage the problem's unique structure: a point on the Pareto front can be generated directly by maximizing accuracy for a fixed privacy level. Motivated by this property, we first derive the shape of the trade-off theoretically, which allows us to model the Pareto front directly and efficiently. To address inefficiency in preference learning, we replace pairwise comparisons with a more informative interaction. In particular, we present the user with hypothetical trade-off curves and ask them to pick their preferred trade-off. Our experiments on differentially private logistic regression and deep transfer learning across six real-world datasets show that our method converges to the optimal privacy-accuracy trade-off with significantly less computational cost and user interaction than baselines.

We statistically analyze empirical plug-in estimators for unbalanced optimal transport (UOT) formalisms, focusing on the Kantorovich-Rubinstein distance, between general intensity measures based on observations from spatio-temporal point processes. Specifically, we model the observations by two weakly time-stationary point processes with spatial intensity measures $μ$ and $ν$ over the expanding window $(0,t]$ as $t$ increases to infinity, and establish sharp convergence rates of the empirical UOT in terms of the intrinsic dimensions of the measures. We assume a sub-quadratic temporal growth condition of the variance of the process, which allows for a wide range of temporal dependencies. As the growth approaches quadratic, the convergence rate becomes slower. This variance assumption is related to the time-reduced factorial covariance measure, and we exemplify its validity for various point processes, including the Poisson cluster, Hawkes, Neyman-Scott, and log-Gaussian Cox processes. Complementary to our upper bounds, we also derive matching lower bounds for various spatio-temporal point processes of interest and establish near minimax rate optimality of the empirical Kantorovich-Rubinstein distance.

Current temporal knowledge graph question answering (TKGQA) methods primarily focus on implicit temporal constraints, lacking the capability of handling more complex temporal queries, and struggle with limited reasoning abilities and error propagation in decomposition frameworks. We propose RTQA, a novel framework to address these challenges by enhancing reasoning over TKGs without requiring training. Following recursive thinking, RTQA recursively decomposes questions into sub-problems, solves them bottom-up using LLMs and TKG knowledge, and employs multi-path answer aggregation to improve fault tolerance. RTQA consists of three core components: the Temporal Question Decomposer, the Recursive Solver, and the Answer Aggregator. Experiments on MultiTQ and TimelineKGQA benchmarks demonstrate significant Hits@1 improvements in "Multiple" and "Complex" categories, outperforming state-of-the-art methods. Our code and data are available at https://github.com/zjukg/RTQA.

The effectiveness of large language models (LLMs) to fact-check misinformation remains uncertain, despite their growing use. To this end, we present CANDY, a benchmark designed to systematically evaluate the capabilities and limitations of LLMs in fact-checking Chinese misinformation. Specifically, we curate a carefully annotated dataset of ~20k instances. Our analysis shows that current LLMs exhibit limitations in generating accurate fact-checking conclusions, even when enhanced with chain-of-thought reasoning and few-shot prompting. To understand these limitations, we develop a taxonomy to categorize flawed LLM-generated explanations for their conclusions and identify factual fabrication as the most common failure mode. Although LLMs alone are unreliable for fact-checking, our findings indicate their considerable potential to augment human performance when deployed as assistive tools in scenarios. Our dataset and code can be accessed at https://github.com/SCUNLP/CANDY

The development of Long-CoT reasoning has advanced LLM performance across various tasks, including language understanding, complex problem solving, and code generation. This paradigm enables models to generate intermediate reasoning steps, thereby improving both accuracy and interpretability. However, despite these advancements, a comprehensive understanding of how CoT-based reasoning affects the trustworthiness of language models remains underdeveloped. In this paper, we survey recent work on reasoning models and CoT techniques, focusing on five core dimensions of trustworthy reasoning: truthfulness, safety, robustness, fairness, and privacy. For each aspect, we provide a clear and structured overview of recent studies in chronological order, along with detailed analyses of their methodologies, findings, and limitations. Future research directions are also appended at the end for reference and discussion. Overall, while reasoning techniques hold promise for enhancing model trustworthiness through hallucination mitigation, harmful content detection, and robustness improvement, cutting-edge reasoning models themselves often suffer from comparable or even greater vulnerabilities in safety, robustness, and privacy. By synthesizing these insights, we hope this work serves as a valuable and timely resource for the AI safety community to stay informed on the latest progress in reasoning trustworthiness. A full list of related papers can be found at \href{https://github.com/ybwang119/Awesome-reasoning-safety}{https://github.com/ybwang119/Awesome-reasoning-safety}.

Traffic accidents can be studied to mitigate the risk of further events. Recent advances in machine learning have provided an alternative way to study data associated with traffic accidents. New models achieve good generalization and high predictive power over imbalanced data. In this research, we study neural network-based models on data related to traffic accidents. We begin analyzing relative feature colinearity and unsupervised dimensionality reduction through autoencoders, followed by a dense network. The features are related to traffic accident data and the target is to classify accident severity. Our experiments show cross-validated results of up to 92% accuracy when classifying accident severity using the proposed deep neural network.

Global climate models parameterize a range of atmospheric-oceanic processes like gravity waves, clouds, moist convection, and turbulence that cannot be sufficiently resolved. These subgrid-scale closures for unresolved processes are a leading source of model uncertainty. Here, we present a new approach to developing machine learning parameterizations of small-scale climate processes by fine-tuning a pre-trained AI foundation model (FM). FMs are largely unexplored in climate research. A pre-trained encoder-decoder from a 2.3 billion parameter FM (NASA and IBM Research's Prithvi WxC) -- which contains a latent probabilistic representation of atmospheric evolution -- is fine-tuned (or reused) to create a deep learning parameterization for atmospheric gravity waves (GWs). The parameterization captures GW effects for a coarse-resolution climate model by learning the fluxes from an atmospheric reanalysis with 10 times finer resolution. A comparison of monthly averages and instantaneous evolution with a machine learning model baseline (an Attention U-Net) reveals superior predictive performance of the FM parameterization throughout the atmosphere, even in regions excluded from pre-training. This performance boost is quantified using the Hellinger distance, which is 0.11 for the baseline and 0.06 for the fine-tuned model. Our findings emphasize the versatility and reusability of FMs, which could be used to accomplish a range of atmosphere- and climate-related applications, leading the way for the creation of observations-driven and physically accurate parameterizations for more earth-system processes.