As AI has become a huge industry, to an extent it has lost its way. What is needed to get us back on track to true intelligence? We need agents that learn continually. We need world models and planning. We need knowledge that is high-level and learnable. We need to meta-learn how to generalize. The Oak architecture is one answer to all these needs. It is a model-based RL architecture with three special features: 1) all of its components learn continually, 2) each learned weight has a dedicated step-size parameter that is meta-learned using online cross-validation, and 3) abstractions in state and time are continually created in a five-step progression: Feature Construction, posing a SubTask based on the feature, learning an Option to solve the subtask, learning a Model of the option, and Planning using the option’s model (the FC-STOMP progression). The Oak architecture is rather meaty; in this talk we give an outline and point to the many works, prior and contemporaneous, that are contributing to its overall vision of how superintelligence can arise from an agent’s experience.

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (Test of Time Award) haoqing Ren, Kaiming He, Ross Girshick, Jian Sun

Paper Abstract: State-of-the-art object detection networks depend on region proposal algorithms to hypothesize object locations. Advances like SPPnet and Fast R-CNN have reduced the running time of these detection networks, exposing region proposal computation as a bottleneck. In this work, we introduce a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals. An RPN is a fully-convolutional network that simultaneously predicts object bounds and objectness scores at each position. RPNs are trained end-to-end to generate high-quality region proposals, which are used by Fast R-CNN for detection. With a simple alternating optimization, RPN and Fast R-CNN can be trained to share convolutional features. For the very deep VGG-16 model, our detection system has a frame rate of 5fps (including all steps) on a GPU, while achieving state-of-the-art object detection accuracy on PASCAL VOC 2007 (73.2% mAP) and 2012 (70.4% mAP) using 300 proposals per image. Code is available at https://github.com/ShaoqingRen/faster_rcnn.

Though seemingly opposite, doom and optimism regarding generative AI's spectacular rise both center on AGI or even superintelligence as a pivotal moment. But generative AI operates in a distinct manner from human intelligence, and it’s not a less intelligent human on a chip slowly getting smarter anymore than cars were mere horseless carriages. It must be understood on its own terms. And even if Terminator isn’t coming to kill us or superintelligence isn’t racing to save us, generative AI does bring profound challenges, well-beyond usual worries such as employment effects. Technology facilitates progress by transforming the difficult into easy, the rare into ubiquitous, the scarce into abundant, the manual into automated, and the artisan into mass-produced. While potentially positive long-term, these inversions are extremely destabilizing during the transition, shattering the correlations and assumptions of our social order that relied on superseded difficulties as mechanisms of proof, filtering, sorting and signaling. For example, while few would dispute the value of the printing press or books, their introduction led to such destructive upheaval that the resulting religious wars caused proportionally more deaths than all other major wars and pandemics since combined. Historically, a new technology's revolutionary impact comes from making what's already possible and desired cheap, easy, fast, and large-scale, not from outdated or ill-fitting benchmarks that technologists tend to focus on. As such, Artificial Good-Enough Intelligence can unleash chaos and destruction long before, or if ever, AGI is reached. Existing AI is good enough to blur or pulverize our existing mechanisms of proof of accuracy, effort, veracity, authenticity, sincerity, and even humanity. The tumult from such a transition will require extensive technological, regulatory, and societal effort to counter. But the first step to getting started is having the right nightmares.

Join us for the inaugural NeurIPS 2025 Startup Pitch Competition, a highlight of our new Mexico City program. This session will showcase a select group of emerging ventures applying novel machine learning models and data-centric approaches to address critical global challenges. Founders will present their core technical innovations and application domains, spanning pediatric neuro-rehabilitation, decentralized data architectures for AI, on-premise industrial intelligence, ML-driven climate monitoring, quantum-based sensing for AI perception, and agentic models for enterprise finance. This session provides a unique opportunity to bridge the gap between foundational research and real-world deployment, highlighting novel applications of machine learning in diverse and high-impact sectors.

Featured Startups: PhantasiAI: Redefining pediatric rehabilitation with an AI stack that powers non-invasive neurostimulation to restore gait in paralyzed children.

PublicAI: Building a decentralized "Human Layer of AI," enabling millions of global contributors to own and monetize real-world data for training next-generation models.

Pluma: Deploying local, on-premise AI agents for industrial SMEs, ensuring data privacy while automating complex engineering tasks on existing factory hardware.

SEKHEM-ETHOS: Using machine learning and satellite imagery for real-time environmental monitoring, bushfire detection, and transparent carbon credit verification in Africa and the Global South.

OAQ: Developing quantum sensors to capture novel magnetic anomaly data, providing a new perception layer to unlock the next era of AI in autonomy, defense, and AGI.

DAKO Labs: Building the first Agentic Enterprise Finance advisor for CFO teams, creating a research-driven path toward autonomous and trustworthy finance operations.

The “Women in AI Social – Amplifying Voices in AI” is an interactive community event designed to foster meaningful connections, peer dialogue, and collective reflection among women and allies in the AI research ecosys tem. The event embraces inclusive, participatory formats, such as rotating roundtables, speed networking, and creative group activities, that center the voices of all attendees. With a focus on mentorship, inclusion, and global-local per spectives (especially from Latin America), the event invites participants to share experiences, build supportive networks, and envision a more equitable AI future. It is open to all registered NeurIPS CDMX participants, regardless of gender or career stage, and culminates in an optional post-event dinner or social outing to continue the conversations in an informal setting.