Tracked across scientific fields
A chronological view of public research related to this topic.
Related updates
How I used AI to investigate millions of historical records and surfaced a forgotten meteorite report, three lost rhinos, and unrecorded volcano eruptions. Last week, on October 1st, 2026, the historian Benjamin Breen published a post called Using Opus 5.5 to discover a new eyewitness account of the dodo. He used AI to search digitized records from the Dutch East India Company. The Company dominated the spice trade a
Ai2 at COLM 2026: Open research, from models to agents for science October 9, 2026 Ai2 This week Ai2 was at COLM 2026 in San Francisco. The work we shared reflects our wider mission: building AI research artifacts anyone can study and extend, and developing AI for science tools in collaboration with domain experts. Our evening gathering gave colleagues from across the COLM and wider open-source community another sett
It’s hard to believe that Periodic was only launched last September : One year later, it is considered one of the pre-eminent AI scientist labs, with dizzying talent density and astonishing progress in the autonomous lab buildout: Most people are familiar with the standard credentials of Liam and Dogus , but we found an incredible “talent slope” while learning more about Periodic, where each successive employee seems
Drug discovery for neurological disease has traditionally centered on the molecules altered by disease. But the molecules that cause pathology are not necessarily the best points from which to reverse it. Here, we ask which otherwise unaltered molecular control points can be engaged to restore pathological neural circuits toward functional states. We present CircuitATLAS, a provenance-grounded systems-neuroscience kn
Humans perceive far more in a scene than what is explicitly depicted: a single glance captures past causes and future trajectories; a quick peek determines if a vehicle can fit between two parked cars; a few seconds of video reveals who holds authority in a room; and a fleeting clip highlights subtle abstract patterns like unwritten rules or hidden labels. This capacity reflects a form of humanity's sixth sense: an i
As AI agents’ ability to do science autonomously continues to improve, scientists are facing a once-unthinkable question: will I get scooped by a bot? At least twice in the past five weeks, researchers have said that they had been working for an extended period on a particular research question, only to learn that artificial-intelligence companies had either answered the question before them or announced results befo
Black-box optimization (BBO) arises in many scientific and engineering problems where objective evaluations are expensive and limited. Recent large language model (LLM) agents offer a new way to approach BBO by combining task semantics, computation, optimization tools, and feedback-driven decision making, showing great potential due to the integration with mathematically rigorous tools. However, existing agentic BBO
Invasive candidiasis has become more common in recent decades, particularly within ICUs. Identifying specific Candida spp., and testing their sensitivity to antifungal drugs is crucial for effective treatment that helps healthcare providers to detect potential treatment challenges early. This study aimed to identify Candida spp. at their spp. level present in blood, assess their susceptibility to antifungal agents, a
Automated building design must comply with seismic and wind codes and satisfy structural mechanics constraints, yet most existing agents produce visually plausible models without verification grounded in mechanical analysis and code compliance. We introduce the Vibe Building task and propose PE-Loop (Physics-Engine-in-the-Loop), an agent in which a deterministic physics engine is the sole source of evaluation signals
Large language model agents can improve across tasks by retaining reusable skills distilled from prior interactions. Recent work jointly optimizes task execution and skill extraction, enabling the policy and skillbank to co-evolve. However, as the actor continues learning, rewarding skill proposals through their reuse in subsequent training steps may conflate skill benefits with actor improvement, while directly test
This work examines the transfer of a co-evolved communication mechanism between two robotic agents from a discrete two-dimensional (2D) simulator to a three-dimensional simulator with real physics (3D). The study focuses on whether a communication mechanism co-evolved in a 2D environment retains its functional role after transfer to a 3D physics-based simulator. To support this analysis, the effects of the episode ti
Dual-process agents pair a fast policy with a slow deliberative model. In real-time settings the slow model usually runs continuously; in turn-based agents and robot planners it is invoked on events such as uncertainty or a detected failure. We study a fast learned actor that takes every decision and hands control to a reasoning vision-language model only when a gate opens, while the game keeps running. We use closed
Latent world models have shown a remarkable ability to predict future states and to plan in the real world. In practice, however, we lack a principled way to estimate how their capabilities scale with model size, data, and compute, an open problem that slows progress in the field. In this work we present RoboJEPA, a world model based on the Joint Embedding Predictive Architecture (JEPA) and trained on a large-scale d
The floodgates are open. Right on the heels of OpenAI’s new large language model producing the biggest math breakthrough in two decades, the company just released no less than 372 results from the same model. Each resolves or makes substantial progress on a major open question in mathematics or theoretical computer science, the company says. OpenAI revealed the results in a GitHub repository at 6 P.M. EDT. The deluge
At the International Conference on Machine Learning (ICML 2026) , Dezhi Luo, Yijiang Li, Maijunxian Wang, Tianwei Zhao, Bingyang Wang, Siheng Wang, Pinyuan Feng, Pooyan Rahmanzadehgervi, Ziqiao Ma , and Hokin Deng presented their work Vision Language Models Cannot Reason About Physical Transformation . In this interview, William Yijiang Li tells us more about the research, the method the team used, and the controlled
Enterprise AI agents that share a memory store face two unaddressed risks: sensitive data can leak through legitimately computed results the requester could not derive, and departments can silently compute a same-named key performance indicator (KPI) through conflicting logic. Existing agent-memory systems (e.g., MemGPT, Zep, A-MEM) gate retrieval by content, ownership, and role, not derivation, missing a cached insi
Modern information systems, including many agentic workflows, use dense retrieval to explore large amounts of unstructured data. However, dense retrieval relies on surface-level semantic similarity, which is insufficient for increasingly complex search applications. Here, we investigate agentic retrieval that combines the reasoning capabilities of Large Language Models (LLMs) with the efficient corpus exploration of
A large language model (LLM) agent solves long-horizon tasks through many reasoning-action turns, with one verification signal at termination. Deployed agents face streams of related tasks, making their trajectories a natural resource for improvement. In-context adaptation agents store reflections, memories, or skills as text, so reuse depends on retrieving the right experience and on a frozen policy executing it. We
LLM-based agents are increasingly capable of generating complex 3D structures, with the potential to reshape how objects are designed and realized in the physical world. Yet, producing elegant geometry is fundamentally different from producing objects that can be built and perform their intended functions. Existing evaluations largely focus on geometric quality while overlooking physical realizability. We introduce L
Language models can already help researchers search the literature, synthesize evidence, and work through complex questions. But scientific work places particular demands on these models—answers need to stay grounded in evidence, the models need to preserve what the evidence actually supports rather than quietly broadening a study’s conclusions, and researchers need to be able to verify the final outputs. We see that
Terminal agents are increasingly deployed beyond software engineering in science and other specialized domains. Constructing training environments requires executable reference behavior and a domain-specific verifier that distinguishes semantic correctness from superficially plausible artifacts. Authoring these components for each task requires repeated engineering and limits reuse. We introduce software-in-the-loop
Reasoning traces improve large language models (LLMs), but current models are trained to reason mostly in English. It has been shown that forcing a model to reason in another language degrades accuracy, even when the reasoning language matches the language of the prompt -- but only for a setting where the model reasons over a short prompt. Here, we ask whether the same holds for retrieval-augmented generation (RAG),
Science Summary: In this guest post, Prof. Matthew Schwartz returns to describe a new approach to AI-accelerated science. In Vibe Physics, Schwartz discussed similarities in capability between Claude and a physics graduate student. Here, he describes what happened when he stopped fighting Claude and allowed Claude to find “Claude-shaped” problems: ones best suited to the capabilities of the current generation of LLM
Scientific progress emerges from a longitudinal ecosystem in which researchers, institutions, funding agencies, collaboration networks, and the scientific literature co-evolve. As AI becomes increasingly involved throughout the scientific research cycle, understanding these interconnected and evolving processes becomes increasingly important. We introduce SciUtopia, a persistent, closed-loop LLM-agent simulation fram
Computer-use agents need to reliably ground action targets in complex desktop scenes, where multiple applications, overlapping windows, and visually similar controls compete for attention. Existing training data rarely pair such scenes with dense annotations or vary them in a controlled way. We introduce DeskForge, a controllable desktop environment that composes and explores real applications to generate large-scale
What makes great scientists great? Even as AI systems start to make progress on open problems, scientists remain far ahead of them at sensing which prior idea, buried in an ever-growing archive of research, a new problem needs. To study this skill, we draw on researchers who know firsthand which earlier work advanced their completed projects, with papers serving as pointers to the ideas within. Using our automated pi
As LLM agents undertake increasingly complex, long-horizon tasks, verifying their outputs becomes increasingly challenging. We study how verification capability can be strengthened with a fixed base model, without access to reference answers or grading rubrics at test time. Repeated sampling yields multiple rollouts that can contain complementary correct claims, but we need a reliable verification mechanism to determ
Noam Brown, an OpenAI researcher and core author of o1, discussed his current focus, multi-agent systems, in a recent podcast. He said OpenAI previously used 10,000 agents over 88 hours and 130 billion tokens to solve a Millennium Prize mathematics problem. Brown was explicit that the result came from having a general, powerful model, and that he would not even attribute 10% of the credit to multi-agent methods, whic
Scientific software presents a demanding test for computer-using agents based on visual language models (VLMs): completing a research workflow requires interpreting specialized interfaces, manipulating scientific objects, and producing verifiable results. We thus introduce OSWorld-Science, a benchmark and evaluation environment that combines scientifically meaningful tasks, artifact-based evaluation, and an efficient
LLM agents generate intermediate reasoning and actions token by token, making extended interactions slow and computationally expensive. Jev-style models offer fast probabilistic predictions over finite fields, but require those fields to be specified in advance. This requirement limits autonomous task solving, where the available actions must be derived from natural language instructions and adapted through interacti
This work evaluates the direct transfer of a co-evolved communication protocol from a 2D simulation to a 3D physical environment, without retraining the network weights. Two e-puck-type robots, controlled by a GRU network with residual connection, were evaluated in a food-seeking task with social signaling. The sensory and motor translation layer required three corrections for stable physical operation, including the
Manipulation behaviors vary widely across objects and scenes, but they share a small set of reusable skills, and planning with these skills helps embodied agents generalize to new tasks. Yet an agent can only plan with skills it knows. Recovering skills from observed experience, the inverse of planning, builds this knowledge over time and yields skill data for training future agents. Vision-Language Models (VLMs) des
Large language models (LLMs) are increasingly deployed as agents for multi-step decision-making, yet transfer poorly to unseen environments. World-model methods address this by training agents to predict future observations, at the cost of additional training and errors that compound when predictions are used for planning. However, for LLM agents operating in digital environments, much of this world knowledge is alre
Coding agents are beginning to move beyond purely digital tasks to tackle physical-world challenges, particularly in robotics. Existing robotics benchmarks, however, primarily focus on the performance of individual artifacts, such as policies or controllers, offering limited coverage of coding agents' broader engineering capabilities. Real-world robotics extends beyond control: agents must build, integrate, diagnose,
Reinforcement learning has helped AI agents solve increasingly difficult tasks, but high rewards do not always reflect the work users intended. In recent incidents and controlled evaluations across the AI industry, agents trained to maximize reward have accessed unauthorized information, attempted to evade monitoring systems, and even breached sandbox protections to attack external systems. As agents become more capa
Microrobots hold considerable potential for various applications in which targeted navigation is a basic requirement. Deep reinforcement learning has recently emerged as a powerful paradigm for fully autonomous microrobot navigation. Yet, current deep reinforcement learning-based approaches pay limited attention to learning efficiency and effectiveness, requiring hours to days for model training. Consequently, this i
Before invoking external tools, an agentic LLM must select among a K-way action space: executing a call, seeking clarification, answering directly, or declining. While internal activation steering can alter these pre-execution decisions, conventional aggregate metrics obscure where altered states land and what collateral damage they inflict. We present SAKIKO, an auditing framework that formalizes representation repa
In response to the growing demand for long sequences in agentic and reasoning use cases, many state-of-the-art LLMs combine multiple variants of attention to mitigate the quadratic complexity of traditional softmax attention. These hybrid attention LLMs aim to balance the strengths and limitations of full attention and alternatives based on recurrence. This work presents a first study of how hybrid attention impacts
Multi-step agents are trained on flat action streams: SFT and RLVR weight every token uniformly and ignore the sub-procedures that recur across tasks, the hierarchy that lets humans plan top-down from reusable routines. This structure sits unused, and flat training uses each scarce trajectory less fully than its content allows. Recent agents do use that structure, but only as LLM-written skills in context, never in t
AI Models & Platforms Lila Sciences’ AI Lab Uncovers Palladium Catalysts for Green Hydrogen Add Unite.AI to your preferred sources on Google Lila Sciences said on September 25, 2026, that its AI-directed laboratory proposed, synthesized and screened 2,942 catalysts for green hydrogen in three months, identifying six palladium-based material families for the acidic oxygen evolution reaction at the heart of hydrogen pr
First up on the podcast, Staff Reporter Tom Howarth joins host Sarah Crespi to talk about the language artificial intelligence agents evolve inside their swarms . Next on the show, producer Elah Feder talks with Revathe Thillaikumar , a behavioral ecologist and postdoctoral fellow at the Max Planck Institute of Animal Behavior. The two discuss the value of social learning as baby orangutans learn to find their own fo
Anthropic touts AI-led biology discovery Andrew Zinin Chief Editor U.S. artificial intelligence startup Anthropic on Wednesday unveiled the first discovery from its molecular biology lab, a previously unknown enzyme system spotted by its Claude model and touted by the company as proof that AI can speed up basic research. Enzyme systems found in nature have led to some of biology's biggest breakthroughs, from the tool
Anthropic touts AI-led biology discovery - Anthropic's AI model Claude discovered a new enzyme system called ART by analysing DNA sequences, potentially revealing a new gene-editing mechanism similar to CRISPR. - The AI autonomously analysed over 200,000 genes in 21 hours, accelerating research that usually takes weeks or months. - The discovery, still preliminary and unreviewed, could lead to advances in gene therap
[This is a guest post by Tapio Schneider . This blog post was initially written in a different file format and converted using AI. — T.] [This will be cross-posted on the CliMA blog .] The apparent proof of finite-time blow-up of the forced Navier-Stokes equation, announced by OpenAI on September 8, has brought into focus a debate about the role of AI in mathematics. The proof was produced with a system of some 10,00
Science Sep 23, 2026 We’re introducing a new life sciences research group and laboratory at Anthropic. Our focus is on fundamental biology research using Claude: exploring datasets of DNA to identify uncharacterized protein families, generating hypotheses at scale, and testing them through experiments in the lab. This post introduces the team behind this work and shares early results in which Claude discovered a nove
A collaborative system of large language model agents brings protein engineering closer to natural-language interaction, lowering the barrier between human intent and biomolecular design. This is a preview of subscription content, access via your institution Access options Access Nature and 54 other Nature Portfolio journals Get Nature+, our best-value online-access subscription 27,99 € / 30 days cancel any time Subs
Nature Computational Science, Published online: 22 September 2026; doi:10.1038/s43588-026-01051-4 Embodied intelligence has enabled machines to act autonomously, yet intelligence remains largely confined to digital models, leaving robotic bodies as passive executors. In extreme environments, such as the human brain, where tissues are mechanically and thermally sensitive, conventional electronics can become limiting.
AI agents are beginning to automate research and development across the AI stack, from improving training efficiency to optimizing inference. A natural next step is to improve the research efficiency of the agents themselves. When an AI research agent's own code is the object of optimization, each accepted rewrite becomes the agent that the next round edits. We refer to this loop as recursive self-improvement. Its si
Tool-calling LLM agents are increasingly deployed in enterprise applications. However, effective evaluation and optimization require high-quality, diverse task datasets that are often difficult to obtain due to privacy and other constraints. Existing synthetic task generation methods often produce generic tasks that ignore an agent's underlying state or database and fail to reflect real-world usage diversity. We prop
LLM agents are increasingly deployed in collaborative settings, yet long-term interaction may give rise to undesirable coordination. We study the emergence of collusion in a long-horizon multi-agent environment: two agents repeatedly complete individual tasks, share task logs, verify each other's work, and receive rewards. We introduce realistic constraints that make compliance with the verification protocol incompat
Skills can improve the performance of Large Language Model (LLM) agents by providing task-specific procedural guidance, while skill optimization further improves their effectiveness through iterative refinement. However, existing skill optimization methods typically represent skills as unstructured natural-language instructions, creating two key challenges: 1) Unstructured skills often lack explicit workflow-level gu
AI Models & Platforms Insilico Medicine Releases Open Longevity AI Toolkit in Cell Study Add Unite.AI to your preferred sources on Google Insilico Medicine on September 17, 2026 announced the publication of a study in Cell introducing an openly released AI toolkit for aging biology: the LongevityBench benchmark, a family of five compact Longevity-LLMs, and the Longevity Claw agentic research platform. Publication in
The paper introduces three interconnected resources for the global scientific community: - LongevityBench, the first open benchmark designed to rigorously evaluate AI reasoning across multiple domains of aging biology; - Longevity-LLMs, a family of compact, open-source language models specifically trained on clinical and multi-omics aging data; and - Longevity Claw, an open-source agentic research platform that auton
ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments 摘要 Scientific code repositories encode decades of human knowledge in executable models, methods, and tools. Yet fragmented toolchains, implicit domain conventions, and specialized correctness criteria make this knowledge difficult to convert into reliable learning experience—a challenge we call the scientific experience bottleneck. We i
It’s a pharmaceutical-company executive’s dream: tens of thousands of employees working day and night, uninterrupted by meals, sleep or distraction, to find the next blockbuster drug. A sophisticated artificial-intelligence system called the Virtual Biotech, which is described today in the journal Science1, is a step towards this vision. The Virtual Biotech comprises as many as 37,000 agents — AI systems that autonom
Understanding why people age — and how to slow that process down — is an obsession that has stretched across millennia, from mythic quests for immortality to modern science’s hunt for drugs that lengthen lifespan. Now, that quest has entered the AI era. A report1 published today in Cell unveils a bespoke artificial-intelligence system that aims to turbocharge longevity research. The system introduces large language m
scBaseCount is presently the largest public single-cell RNA-seq repository, containing over 502 million cells across 27 organisms and 75 tissues. An AI agent autonomously discovers, annotates, and uniformly reprocesses all 10× Genomics datasets in the SRA, creating a harmonized, continually updated resource for studying the diversity of cell biology and training AI models.
Enabling robots to adapt to unfamiliar environments as readily as humans remains a moonshot goal of embodied AI. No finite collection of demonstrations can cover every task and situation a robot will encounter, making the ability to learn from context at deployment essential for generalization. Such in-context learning (ICL), however, remains largely beyond the reach of existing robotic policies. The broad agentic ca
ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents 摘要 ScienceBuddy is an interactive scientific research workspace that brings continually improving agents into a researcher's everyday work. It supports researchers in carrying out scientific tasks while turning their requests, feedback and execution evidence into tasks and evaluation rubrics for continual learning. At its core is
On September 8, OpenAI said 10,000 of its agents, running for 88 hours, had produced a 166-page proof regarding the Navier-Stokes equations — which describe fluid movement and have remained an open question since 1934. It's one of the seven Millennium Prize Problems established by the Clay Mathematics Institute in 2000, each carrying a $1 million prize; only one has ever been awarded. OpenAI has said it won't be clai
A statement signed by 25 Fields medalists, posted by Terence Tao on September 11 under the title 《人工智能在数学中的严重错位》, argues that AI companies' focus on solving famous mathematical problems seriously diverges from the community's goals of conceptual understanding, method development, and knowledge transmission. Former OpenAI VP of post-training Liam Fedus responds that today's AI limits should not frame future possibilit
Scientific research is a continuous process that emphasizes inheritance. Methods developed by predecessors are often expanded upon by new researchers to explore more novel and in-depth scientific questions. However, the change of lab staff, such as student graduation, leads to a lack of personnel capable of replicating methods. Methods that have been developed with significant effort and resources cannot be continued
Large Language Model (LLM) agents offer a promising path toward autonomously managing long-term physical tasks without human intervention. However, physical tasks require agents to continuously observe the environment, make consequential actions, and remain effective as the environment changes. Existing approaches either require substantial data and retraining, or primarily focus on agents operating in the virtual wo
LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in such consultation settings remains challenging since (i) it requires setups where the assistant learns about social situations from subjective user narratives, and (ii) social properties, such as others' intentions, typically lack verifiable ground truth. To address these challenges, we introduce Fuse, a multi-agent simul
PhAI Labs announced its DFM research framework, published the DFM technical report and research-map repository, and said ScienceBuddy, ScienceIDE and JEPA-Anything were scheduled as separate releases on September 16–18.
Discovery Foundation Models: Toward Open-Ended Discovery Intelligence 摘要 Foundation models have progressed from learning and reasoning over existing knowledge, to increasingly learning through action, tool use, and outcome feedback. We argue that the next frontier is a further transition: from solving and acting within problems specified by humans to participating in the process by which new problems, representations
PhAI Labs opened a scientist collaboration program focused first on life science and biomedicine, inviting collaborators with scientific questions, data, computational or experimental environments, and real validation conditions.
Machine learning, at its core, is about generalization, not memorization. You hand your learning algorithm a pile of training examples and use them to fit a model. But the goal is not to perform well on the training examples — that's easy, you could just memorize the answers. The goal is to perform well on new examples that you have never before seen. If a model does well on the data it was trained on but poorly on f
AutoDiscovery, our AI agent for scientific research, analyzes datasets, proposes hypotheses, runs experiments to test them, and ranks the results by Bayesian surprise—the gap between what the model predicted and what the data showed. In a University of Washington classroom this spring, it surfaced a possible flaw in how battery aging is measured, an unexpected pattern in silicon polymers, and a recurring feature in p
Backdoor poisoning attacks add poisoned examples to otherwise-clean finetuning data, pairing a trigger with a target behavior that the model learns to produce when the trigger appears. Existing evaluations typically fix the number of poisoned examples and sample them at random from a candidate pool. We show that this can severely underestimate worst-case vulnerability: across three LLaMA-3-8B backdoor settings, holdi
Digital agents must often adapt to new environments whose interfaces, tools, and failure modes are not fully captured by pretrained models. We introduce RSIAgent, a training-free multi-agent framework for recursive self-improvement through autonomous memory construction. RSIAgent coordinates curriculum, actor, and verifier agents to continually explore the environment, validate outcomes, and retain environment-specif
When people engage in spontaneous thought under low external constraint, thought content changes continuously and can shift across topics. However, the cognitive computational principles governing these transitions remain unclear, as does whether similar dynamics emerge in artificial language generation. Here, we formalized spontaneous narrative dynamics using a foraging-inspired framework, constructing an agent base
It was a week that left mathematicians reeling. Hot on the heels of a flurry of cases of artificial intelligence furthering the field, OpenAI declared a major scalp: its latest AI model had cracked a Millennium Prize Problem, a puzzle with a $1m reward that had defied human brains for decades. The achievement bore little resemblance to how mathematical problems normally fall. A near-trillion dollar private company ha
Twenty-five Fields Medalists, with Terence Tao and Deng Yu among the first signers, issued an urgent public statement arguing that the way AI companies are pursuing mathematics is now in "severe misalignment" with the goals of the mathematical community. The statement explicitly does not call for stopping AI or deny its achievements. It raises three concerns: AI is solving problems faster than the community can diges
The controversy over OpenAI's maths miracle ‘Historic’ mathematical breakthrough by the tech company has been overshadowed by data privacy concerns of human researchers This year has been one of “unthinkable” progress for maths, said The Wall Street Journal. But OpenAI’s proof of the Navier-Stokes problem – one of the seven esteemed Millennium Prize Problems accepted as maths’ “deepest and most difficult unresolved q
We demonstrate that a general-purpose agent can directly drive a physical robot throughout task execution without any task-specific or environment-specific training. We introduce Agent as Policy (AGP), which places task planning and execution under the agent's control. Given a task and a robot interface, the agent interprets visual evidence, writes executable programs, issues motion commands, and revises its actions
Nature Machine Intelligence, Published online: 10 September 2026; doi:10.1038/s42256-026-01298-6 Shi et al. demonstrate a dual architecture approach for materials research, integrating two lightweight large language models for collaborative reasoning and scientific tool execution. The method achieves competitive performance while remaining affordable and locally deployable.
Nature Computational Science, Published online: 10 September 2026; doi:10.1038/s43588-026-01049-y MutexaGPT turns plain-language enzyme-engineering intuitions into lead mutation designs through AI agent-orchestrated high-throughput molecular modeling, yielding experimentally validated improvements in enzyme specificity and cold activity.
Neurosymbolic systems rely on mathematical solvers to guarantee reasoning correctness, yet solvers are fundamentally blind to whether a formal translation maintains strict reference-equivalence to a designated formalization. We formalize this vulnerability as Verdict-Preserving-Unfaithfulness (VPU): a failure mode where an incorrect encoding executes successfully and matches the expected verdict. We theoretically pro
Today was a tough news cycle to launch anything; we ordinarily promise to cover any new decacorn fundraises so Cognition’s $48B round and Mistral’s $24B round would normally have made it; we love imagegen so GPT Image 2.5 would have been its own headline; we covered the Dreamer story closely so their relaunch as Meta’s Muse agent should have made it; but.. yknow… the bar is higher these days. The summaries below capt
Reacting to sudden physical hazards (catching a slipping plate, dodging a falling knife) is both a meaningful test of embodied intelligence and a hard requirement for deploying multimodal large language models (MLLMs) as the decision coreof household robots. Existing evaluations, however, probe intuitive physics passively through question answering over videos, or target deliberate, long-horizon tasks such as navigat
A research idea may be novel, coherent, and scientifically plausible, yet its proposed method may remain insufficiently specified for faithful implementation. We study the codification readiness of implementation-facing research-method specifications, defined by whether they provide sufficient methodological information for a competent implementer or coding agent to construct the intended method without unsupported a
There’s some exciting very recent work by Alpöge and Buckmaster , building upon prior work by Córdoba and Martínez-Zoroa , in the general topic around the infamous global regularity problem for the incompressible three-dimensional Navier-Stokes equations . It is now widely expected that it should be possible to construct smooth initial data and smooth forcing term that would make these equations develop singularities
Self-driving laboratories (SDLs) combine automated experimentation with adaptive decision-making to accelerate scientific discovery. Their operation nevertheless often depends on human specialists who translate scientific objectives into executable closed-loop campaigns. Specialists adjust them as data and operating conditions change. Here, we present La Agente \'Optima, an agentic framework that constructs and super
Across application domains, predictive signals often sit in unstructured free text rather than structured fields, yet turning that text into useful and interpretable features is difficult. Running large language models (LLMs) over an entire corpus is costly and hard to reproduce, while end-to-end text representations can rely on surface cues that are difficult to inspect. We present an agentic workflow that takes a p
Science Sep 4, 2026 We are sharing the first complete computer-checked proof of Fermat’s Last Theorem. Claude worked largely autonomously over 11 days to write the proof in the Lean programming language. Below, we describe how the formalization was done and share some thoughts about what this work could mean for research mathematics. Around 1637, Pierre de Fermat jotted down a claim in the margin of his copy of Dioph
Digital twin technologies could transform neuroscience and biomedicine by creating predictive computational representations of living organisms. However, most animal digital twins model neural circuits, anatomy, or biomechanics separately rather than integrating the processes that generate behavior. We argue that animal digital twins should instead be conceived as embodied dynamical systems that unify neural activity
Black-box optimization problems remain challenging because of large, weakly structured, and high-dimensional search spaces. Existing methods often suffer from poor sample efficiency because they rely on direct candidate generation or trial-and-error refinement. A natural way to improve search efficiency is to use world modeling, which can help identify promising optimization directions before costly evaluation. Large
Scientific software is often locked in its source language, callable from others but not natively extensible or built upon. We present Bio-Babel (biobabel.stanford.edu), an autonomous framework that rebuilds software natively in target ecosystems and makes it agent-callable, shipping each package with an agent-readable contract of its usage. Reconstructing a hierarchical, interdependent R-to-Python single-cell stack,
Powder X-ray diffraction (XRD) is central to materials characterization, yet reliable end-to-end automation remains challenging. An XRD agent must interpret diffraction evidence, operate refinement software, manage coupled parameters in a defensible order, and distinguish numerical improvement from physical validity. In this paper, we propose AutoXRD, an autonomous large language model (LLM) agent framework that orga
Brain-computer interfaces translate neural activity into device commands, yet their performance hinges on preprocessing that remains manual, expert-dependent and poorly reproducible. Large language model agents can automate scientific coding, but existing systems lack the modality coverage, raw-data isolation, experience accumulation and domain oversight that neural preprocessing requires. We introduce EasyBCI, a two
Insect pollination is critical for global food production, yet monitoring pollinators at commercial farm scale remains a challenge. Recent advances in computer vision and deep learning have enabled detailed analysis of pollinator behaviour, but monitoring must trade-off detail against spatial coverage and human or technological resources. This paper presents the Automated Guided Robot for Insect and Crop Activity Mon
Scientific progress depends not only on finding solutions, but on learning the rules that explain why they work and using that understanding to design better experiments. We introduce science sandboxes, a framework for studying this capability in AI agents through repeated cycles of experimentation, feedback, and hypothesis revision. Science sandboxes invite an agent to query the natural world in different ways, rang
Large Language Model (LLM) agents have shown promise for automating data-science workflows, yet their end-to-end performance depends critically on the agent harness that represents tasks, manages execution state, constrains output artifacts, and provides evaluation feedback. Existing data-science agents often leave this harness implicit, making results difficult to reproduce, compare, and attribute across heterogeneo
Protein co-abundance clusters preserved across tissues can reveal shared disease mechanisms and candidate therapeutic targets, particularly when proteins implicated in organ-confined diseases converge in peripheral or accessible tissues. However, previous cross-tissue studies have focused on biologically pre-selected tissue pairs, leaving most possible combinations and non-obvious relationships unexplored. We present
Large language model (LLM) agents are increasingly embedded in scientific workflows for literature analysis, drafting, and review. Existing systems advance autonomous discovery and manuscript generation, but do not resolve the governance problem that arises when ideas, methods, results, and claims propagate through AI-assisted workflows without mandatory human approval or artifact-level traceability. This paper propo
This work provides an overview of the different strategies that can be used to evaluate the performance of AI models and agents based on large language models (LLMs) for materials synthesis. After providing a brief overview of the key technologies behind the current generation of AI agents based on LLMs, we summarize the different approaches to evaluating these models in the context of materials science and in partic
Scientific agents can produce plausible answers while remaining unable to establish whether the computation behind an answer is executable, recoverable, or reproducible. We present BloClaw, an AI4S workstation built around a simple principle: a scientific agent should know what it can do, show how it did it, and state what remains unvalidated. Each capability declares an execution state, input constraints, dependenci
AI agents for therapeutic reasoning across biological contexts - Michelle M. Li, Carnegie Mellon University - Microsoft Research New England Generative Modeling & Sampling Seminar Therapeutic hypotheses can transfer across diseases but their relevance depends on biological context. The same target, perturbation, or treatment can produce different effects across cell types, disease states, genetic backgrounds, and pat
Heuristic search underlies planning in autonomous systems ranging from warehouse logistics to robotic navigation, yet generic heuristics fail to exploit the structural constraints that govern constrained spatial domains, causing search performance to degrade catastrophically on harder instances. We study this problem through the two-column Flying Block Puzzle, a rigorously NP-complete spatial planning microworld whos
Graphical user interfaces have emerged as an important environment for evaluating autonomous AI agents on multimodal interactive tasks. Existing benchmarks such as AndroidWorld and MobileWorld provide strong foundations for mobile agent evaluation, but their application coverage and task design do not yet fully capture the diversity and complexity of realistic mobile use. We present GMA, a benchmark for evaluating ge
The widespread adoption of clinical large language models (LLMs) introduces significant risks of automation bias, premature closure, and clinician deskilling. Current interpretability paradigms, including latent space trajectories, Concept Activation Vectors, and Concept Bottleneck Models, suffer from topological stagnation, metric distortion, and epistemic occlusion, frequently masking intermediate diagnostic uncert
As AI begins to build itself, automating alignment research becomes increasingly important to let safety research keep pace. Although measuring the success of alignment research is enormously challenging, researchers (at Anthropic and elsewhere) have developed benchmarks and automated auditing tools, such as Petri, that quantify common alignment failures, like deception, sycophancy, and jailbreaks. In one of our earl
Battery Prognostics and Health Management (BPHM) is critical for ensuring the safe, reliable, and cost-effective operation of batteries across electric vehicles, grid storage, and consumer electronics. Conventional BPHM approaches, including physics-based models and task-centric deep learning methods, face challenges in computational efficiency and parameterization, cross-domain generalization, dependence on extensiv
Accurate background intensity estimation is crucial for precise astrometric measurements in astronomical imaging, particularly in complex scenarios such as those encountered in Cassini Imaging Science Subsystem (ISS) observations of Saturn's ring system. Traditional methods, like polynomial fitting and statistical method, often fail in non-uniform conditions, such as those caused by Saturn's rings or scattered light,
Natural language driven autonomous co-scientist workflows involve a fundamental trade-off between flexibility and reasoning at the expense of determinism, reproducibility, and observability. Such agents increasingly must communicate across institutional boundaries, where federation topology can shape latency and cost. We systematically evaluated these tradeoffs using a controlled ablation on a production agentic plat
Artificial intelligence (AI) promises to accelerate biological research by automating computational analyses. Yet the ability of AI agents to carry out computational biology at the scale of complete research studies has not been systematically evaluated. Here we introduce BixBench3, a benchmark that measures the capacity of AI agents to process raw biological data through to scientific results. We designed BixBench3
Large language models (LLMs) have shown strong capabilities in reasoning, planning, and tool use, but many scientific and engineering tasks require more than plausible text and code generation. They require understanding how a system responds to intervention, which in practice depends on controlled experimentation. In this work, we propose a multi-agent framework that enables LLM agents to conduct controlled experime
The standard genetic code uses 64 codons to encode 20 canonical amino acids across domains of life. New-to-nature genetic codes enable new chemistries, therapeutics and ecosystem engineering, but recoding the genome of an organism is exceptionally challenging 1-6 . Here we describe automated genetic tRNA expansion (AGENTEX) for multiplexed robotic prototyping of genetic codes in cell-free translation systems. Two Wat
Next AI Breakthrough Could Come From Physics, Says Max Welling Max Welling, co-founder of CuspAI, discusses how physics principles could unlock the next AI breakthrough, accelerating material discovery and informing AI architectures. 5 min read Visual TL;DR Max Welling proposes physics principles as the next significant leap for AI From the article 8 mentionsThe conventional wisdom in artificial intelligence research
Literature reviews are essential to scientific progress, but rigorously evaluating automatically generated reviews remains difficult because many aspects of research utility depend on expert judgment rather than reference-overlap metrics. We introduce LitReview Arena, a battle-style evaluation platform with a structured protocol tailored to literature review quality: domain experts with AI paper-writing experience co
The rapid evolution of energy structures has positioned microgrids as pivotal components of next-generation power systems, offering enhanced resilience and renewable energy integration. However, the inherent low inertia, complex dynamics, and poor model conditions of microgrids necessitate advanced data-driven frequency control strategies. Although reinforcement learning (RL) has demonstrated certain potential and ad
Pretrained genome language models provide reusable representations for DNA sequence analysis, but turning them into reliable downstream predictors remains non-trivial. Their practical performance depends strongly on fine-tuning recipes, and default recipes reported in prior studies may be suboptimal for new tasks or model backbones, making weak downstream results difficult to interpret. These requirements place a sub
Advances in large language models (LLMs) have fueled a wave of research into agency: the ability to reason, plan, and act. This effort has produced agentic frameworks that orchestrate perception, memory, and decision-making around powerful LLM backbones. With the advent of large multimodal models (LMMs), these systems can process and integrate diverse modalities, including images, audio, and video, thereby improving
The coordination of multi-scale tasks is an effective strategy for computational materials discovery, yet the repeated application of diverse algorithms and tools renders it challenging. We report MAESTRO, a large language model (LLM) agent system capable of executing the entire screening pipeline for metal-organic frameworks (MOFs). It processes a large body of MOF literature, links relevant publications to their cr
Background. Artificial-intelligence design tools now plan genome-scale edits, and agentic systems execute those plans with progressively less human oversight. Biosecurity controls are limited to two points: refusal guardrails at the foundation model and sequence-identity screening at the synthesiser. The design stage between them, where the plan is specified, remains governed by recommendations rather than any deploy
Objective Translating free-text clinical trial criteria into computable code sets is a valuable standardization practice that is necessary for producing reproducible real-world evidence studies but requires standardized interpretation across multiple clinical vocabularies. Methods We developed TrialCode Agent, a hybrid-large language model (LLM)-terminology verification agent that generates, formats, verifies, and ex
According to Latent Space, since 2022, components of the AI pipeline have been shifting from human-made to model-made, each flip starting with a 'patient zero' at a frontier lab. This trend is characterized as simulation: slightly worse (10%), but 100x cheaper and 10,000x faster. Stage 1 (2022): Reward signal went synthetic with InstructGPT's reward model, followed by Constitutional AI and RLAIF, where AI critiques i
This tutorial introduces AutoFigure, a practical toolkit for generating scientific figures directly from text descriptions, paper-like content, and structured methodological explanations. Steps include setting up the AutoFigure environment, fixing dependency issues such as Pillow compatibility, preparing SVG and PNG rendering tools, building a custom reference figure, configuring an API-backed generation workflow, an
NVIDIA's coding agent AVO achieved a perfect score on ARC-AGI-3's 25 public games, solving all 183 levels. Given no rules or stated goals, the agent learns by trial and error, observing outcomes and correcting mistakes. AVO's success hinges on retaining what it has learned.
The high photon flux at synchrotron beamlines allows for the measurement of fast dynamical processes. However, beamline radiation-safety protocols prohibit human intervention during X-ray experiments, limiting the ability to perform versatile real-time sample manipulations during continuous data acquisition. Here we present a robotic platform at an X-ray scattering beamline to enable real-time sample handling and pro
This survey proposes viewing agent evolution as dynamic graph transformation, modeling agent state as a dynamic graph where memories, tools, skills, workflows, and inter-agent relations are represented as typed nodes, edges, and subgraphs updated through schema-constrained rewrites. The authors organize existing dynamic-graph-based methods for self-evolving agents into four taxonomies: node/feature evolution, edge/to
Nature Computational Science, Published online: 20 August 2026; doi:10.1038/s43588-026-01035-4 As large language models begin to plan and run experiments on their own, the reflex is to demand that they be interpretable before they are trusted. We argue that what grounds trust in autonomous science is not a view inside the model; it is provenance: a complete, re-openable record of what was reasoned, done and measured
Customer-service LLM agents must follow organizational policy when acting on a user's behalf. Compliance failures arise from either forbidden actions, such as granting an ineligible change, or omitted procedural requirements, such as identification or confirmation. Runtime safeguards can intervene on risky actions, but action-local checks do not guide an agent through a multi-step procedure. PolicyGuide compiles each
Existing code generation systems often assume a predefined repository architecture, which fails in zero-to-all code generation where an agent must build a complete software project from natural-language requirements while maintaining modular architecture. We present Repo0, a continuous structural evolution framework that maintains an explicit architectural state as a Dual-Directed-Acyclic-Graph (Dual-DAG), combining
LLM agents learn by interacting with environments, yet existing environments are hand-built and static: blind to an agent's weaknesses, and quickly left behind as it improves. While recent environment generation methods attempt to address this, they require domain-specific pipelines, rely on expensive or unreliable verifiers, and still produce static environments. To alleviate the engineering burden, we propose Envir
What happens when you teach an LLM-based agent the scientific method? Motivation: Scientific discovery emerges from cycles of hypothesis, implementation, empirical testing, and feedback. Can this process be automated? We approach automated algorithm design through the lens of the scientific method, where an LLM-based agent goes through each step of the process in an ordered, iterative fashion. Results: We present The
Training terminal agents requires scalable executable supervision, yet synthesizing high-quality terminal tasks is challenging. Each task couples an instruction, an initialized environment, a reference solution, and an executable verifier; inconsistent assumptions can make tasks unsolvable or mis-evaluated. Multi-stage synthesis may discard original goals, dependencies, state transitions, and procedural constraints.
Bayesian optimization (BO) is a sample-efficient, surrogate-based approach to black-box optimization (BBO), but its evaluation remains dominated by synthetic functions and hyperparameter optimization (HPO) tasks that are typically low-dimensional and single-objective. Engineering design poses a substantially different regime: problems are physics-based, often high-dimensional, constrained by requirements such as cost
Automated structure elucidation from routine spectroscopy remains a major challenge. Researchers introduce MACROS, a multi-agent system that emulates expert iterative hypothesis testing. Trained on 100 million simulated and 1.6 million experimental spectra-molecule pairs, it natively supports arbitrary combinations of routine spectroscopic techniques and achieves unprecedented zero-shot generalization to diverse real
LEGO-RL is a framework that bridges native coding-agent harnesses with scalable policy-gradient optimization without modifying their internal control flow. It ensures faithful optimization via in-process LLM proxying for token-level alignment, reliable execution through scalable sandbox orchestration with image caching and stage-wise defenses against reward hacking, and observable training via an integrated plugin an
Modern agents operate within agent harnesses that manage tools, context, and control flow. The original Agent Lightning proposed a disaggregated architecture connecting arbitrary agents to RL training via an LLM endpoint proxy—a paradigm termed harnessed agentic RL. v1.0 is a lightweight framework of about 3,500 lines of code, supporting arbitrary harnesses and offering a reproducible pipeline for coding-agent RL. Wi
Reinforcement learning (RL) has shown promise in single-turn LLM fine-tuning, but long-horizon agentic reasoning introduces increasingly branching interactions and sparse rewards, exposing limitations: heavyweight backpropagation-based training stacks and harder credit assignment. This paper argues that evolution strategies (ES) are a better choice, offering model scalability (full-parameter optimization with inferen
Florent Delgrange won the Best Blue Sky Paper Award at AAMAS 2026 for his work 'Foundation World Models for Agents that Learn, Verify, and Adapt Reliably Beyond Static Environments.' The paper proposes a research agenda that merges reinforcement learning and reactive synthesis into a single loop: a foundation world model is learned from experience but structured for a verifier to reason about. As the agent learns a p
In mixed traffic, decision-making for autonomous vehicles faces three interrelated challenges: physics-based priors fail to capture latent interactive intentions and diverse driver behaviors; abrupt maneuvers cause non-stationarity, leaving long-tail safety events under-explored; and hybrid action spaces destabilize unified RL training due to temporal scale differences between continuous car-following and discrete la
Post-training papers, model cards, and blog posts often treat scores on a small set of coding benchmarks (e.g., SWE-bench and LiveCodeBench) as evidence of broad coding capability, both for research artifacts and user-facing systems. We argue that optimization for these benchmarks leads to measuring task-specific performance, creating a meaning gap between measured scores and claims of general coding ability. We exam
We present RapidLiDAR, a LiDAR scene completion method that treats initialization as a learned, data-driven component. An adaptive initialization module predicts spatially varying displacement for each partial input point, avoiding manual noise tuning. A multi-scale reconstruction module refines point positions using voxel and BEV feature maps. On SemanticKITTI and KITTI-360, our method achieves state-of-the-art comp
Reliable uncertainty quantification (UQ) is essential for deploying large language model (LLM) agents in complex interactive environments. Existing UQ methods often rely on local signals, such as token probabilities or predictive entropy, overlooking long-range dependencies where errors accumulate across an execution trajectory. We propose RUPA, a trajectory-level UQ framework that represents an execution history as
Google AI Developers demonstrated a workflow that gave Gemini 3.7 Flash hundreds of PDFs from a Victorian-era book of botanical illustrations, extracted the historical plants, classified them with the modern APG IV system, and presented the result as an interactive visualization.
Embodied agents are increasingly used to close the gap left by end-to-end policy models, but existing harnesses remain largely open-loop, reflecting only after episodes, which cannot govern real-time physical interaction. We present Zetta, a closed-loop embodied harness that evolves code-based runtime critics and recovery skills online while keeping the base policy frozen. Through three timescale-separated loops, Zet
Looped language models have shown promising results on reasoning benchmarks, yet their potential for agentic tool use remains largely unexplored. We study this question in compositional tool-calling settings, where models must coordinate multiple API calls, maintain intermediate state, and preserve dependencies across tool interactions. We evaluate native and retrofitted looped language models on API-Bank, BFCL, and
MathCode converts natural-language mathematics problems into Lean 4 theorem statements and attempts formal proofs. Its public project combines a persistent Lean REPL, reusable theorem and axiom libraries, agentic proving, parallel subgoal trees and an Obsidian theorem graph.
Prime Intellect ran 153 internet-isolated autonomous runs across 18 frontier models on the nanoGPT optimizer speedrun, releasing the traces, monitor reports, per-run ledger and harness.
LLM-based agents act on behalf of users to access cloud services, call tools, or invoke agents, but permissions remain static and each request is evaluated independently, allowing deviations from delegated tasks, prohibited combinations of actions, or unrestricted sub-delegation. The Agentic Principal Chain (APC) tracks delegated authority and evaluates requests against accumulated session state using six checks, car
Memory is becoming core infrastructure for long-horizon LLM agents, yet existing evaluations offer limited guidance on which memory substrate—the underlying medium in which memory is represented and stored—should be used under different operating regimes. We present a controlled harness evaluation of memory substrates for memory-augmented agents, covering dense and sparse indices, text records, structural stores, hie
DeepMind’s Pushmeet Kohli on AI’s Scientific Revolution Pushmeet Kohli leads AI for Science at DeepMind, where his team has created AlphaEvolve, an AI system that discovers entirely new algorithms and proves mathematical results that have eluded researchers for decades. From improving 50-year-old matrix multiplication algorithms to generating interpretable code for complex problems like data center scheduling, AlphaE
AI co-scientists are beginning to transform research by generating hypotheses, retrieving related work, designing experiments, executing code, and drafting papers, but current systems ignore the individual researcher, optimizing generic metrics like novelty and reviewer scores. This paper introduces personalized auto-research, which conditions every stage on a representation of the researcher, arguing that personaliz
AutoResearchEval introduces 100 tasks spanning 7 scientific domains and the full research lifecycle, including ideation, retrieval, execution, analysis, writing, and review. Evaluating 8 harness-model combinations yields 800 trajectories with process-level annotation, leading to the AutoResearch Failure Taxonomy (ARFT), a framework of 45 empirically-grounded failure patterns. A human-calibrated agent-as-a-judge pipel
LLM coding agents issue Bash commands through interfaces that may serialize, wrap, and reparse model output. Matched execution scores alone cannot distinguish command-generation errors from failures introduced after generation. QuoteBench measures this boundary with exact final-state validation on 56 one-shot tasks from 14 incident-derived families, crossing the generation contract with the execution transport around
Existing AI scientist systems typically reason over text, code, or precomputed summaries, missing crucial spatial, temporal, and cross-channel evidence. To address this, researchers introduce OmniScientist, an end-to-end omni-modal AI scientist that directly conducts multidisciplinary research from heterogeneous raw evidence. The system features a perception layer and three autonomous agents for ideation, experiment,
This post concerns the following conjecture of Sendov , as well as its strengthening by Phelps–Rodriguez : Conjecture 1 (Sendov’s conjecture) Let , and let be a degree polynomial with all zeroes in the unit disk. Then for every zero of , there exists a critical point of with . Conjecture 2 (Phelps–Rodriguez conjecture) Let , and let be a degree polynomial with all zeroes in the unit disk. Then for every zero of , the
AINews reports on a new paper demonstrating how encrypted reasoning traces from frontier models can be decoded and replayed to other models, sessions, or users, leading to dramatic improvements in open models. Since the release of o1, reasoning models have obscured their traces with cryptographic signatures to prevent distillation, but this study shows that these encrypted blocks can be replayed to weaker models and
Creative agents still lack an effective way to learn from high-quality human films, limiting their ability to produce cinematic-grade videos. To address this, researchers propose the Agentic Video Auto-Encoder (AVA-Encoder), a framework for learning agent-native video representations via agentic auto-encoding. AVA-Encoder transforms a video into a knowledge graph (KG) representation and reconstructs it back into vide
In this crosspost from AI Matters, a publication of ACM SIGAI, Ella Scallan interviews Akari Asai, winner of the 2025 AAAI Doctoral Dissertation Award. Asai's doctoral research focused on retrieval-augmented language models (LMs), arguing for moving beyond monolithic LMs to augmented LMs that are designed, trained, and deployed with complementary modules. She pioneered retrieval-augmented LMs, which retrieve relevant
Johnson et al. report an autonomous three-bar tensegrity robot capable of robust locomotion across varied terrains, maintaining functionality even after extreme impacts, including a 5.7-meter drop onto asphalt. The findings, published online in Nature Machine Intelligence, highlight the potential of tensegrity architecture for enhancing impact resistance and adaptability in robotics.
This work introduces Hierarchical Self-Improvement (HSI), a framework where a single frozen LLM operates across three hierarchical scopes: a task harness, an evolver that rewrites the harness, and a meta-evolver that rewrites the evolver's strategy. The harness is task-specific and continuously evolvable via a fixed injection seam, using environment feedback. A thinking-on/off design isolates the contribution of harn
Ai2 expands collaboration with Hugging Face to accelerate open science August 6, 2026 Ai2 Ai2 and Hugging Face are expanding their partnership to make fully open AI easier to access and deploy. Hugging Face serves as a central platform for the AI community, and the expanded collaboration gives us substantially more storage and bandwidth to support our growing portfolio of models, datasets, benchmarks, and application
CogTwin, presented at IJCAI 2025, is a hybrid cognitive architecture framework designed to bring autonomous reasoning and real-time adaptation to digital twins. It addresses the gap in current digital twins that are largely reactive, relying on pre-programmed rules and static models, and thus struggle with unforeseen events. By integrating symbolic AI, sub-symbolic AI, and neuro-symbolic AI, CogTwin aims to equip dig
Microsoft introduces Echoverse, a suite of twelve training worlds for computer-use agents: ten deep domain worlds and two capability worlds, each drilling a single control (e.g., date pickers, nested filters) in varied forms. Depth is key: these worlds replicate real application behavior, come seeded with realistic data, and maintain coherent state across screens and users. Training on all twelve nearly doubles a 9B
Microsoft Research introduces EvoLib, a framework that enables large language models to learn from their own experience during inference without ground-truth labels or external feedback. EvoLib transforms past attempts into reusable skills and reflective insights, continually refining, consolidating, and reweighting them to turn instance-specific observations into increasingly general knowledge over time. This evolvi
The author recounts attending a workshop on mathematics and AI, which produced the Leiden Declaration, now signed by over 3000 people. Although he did not sign, he clarifies it is not due to disagreement but rather uncertainty about some confident assertions and recommendations. He appreciates the declaration's role in focusing attention on pressing issues, and quotes its passage on characteristic values of mathemati
The air in forests is filled with animal sounds, containing rich information about biodiversity. However, analyzing the massive recordings collected by autonomous devices is a huge challenge. Dan Stowell, a researcher at Naturalis Biodiversity Centre in the Netherlands, leads BioacAI, a four-year EU-funded project with European and UK partners, to develop AI tools that automatically identify species from sound record
A Belief Flow Network (BFN) is a directed graph where nodes represent agents holding epistemic states—complex logical beliefs including conditional statements. Agents exchange information asynchronously, revising their beliefs via iterated belief change operators, potentially with different rational policies (some skeptical, some credulous). The research asks whether consensus is inevitable if influence can eventuall
Accelerating the frontiers of scientific discovery: Google’s $40M commitment to the Genesis Mission Pushmeet Kohli VP for Science and Strategic Initiatives, Google DeepMind, and Chief Scientist, Google Cloud Karthik Narain Chief Product and Business Officer, Google Cloud Google AI for Public Sector Empowering an Adaptive, Responsible, Secure, and Intelligent way forward Learn more Scientists today face challenges of
In April 2026, a NASA Artemis II photo of Earth caught public attention, but AI can fabricate similar images in seconds, raising trust issues. The author, a visual science communication researcher, notes that AI-generated images infiltrate science: used for illustrations, synthetic data, editing lab images, and outreach, blurring lines between illustration, enhancement, and fabrication. In 2024, two papers were retra
“The future of biology is programmable.” To give you a sense of how important this topic is, Patrick Hsu is the first guest we’ve had back to Ground Truths since the podcast began a few years ago. Nearly 2 years ago we spoke about digital biology but so much has happened since to help unravel the complexities in the language of life. Patrick is a co-founder and core investigator at the Arc Institute . His work is at
I believe that the creation of visualization apps to illustrate mathematical or scientific concepts is a particularly favorable use case for modern coding agents, as many of the downside risks attached to other LLM use cases are limited: Not mission-critical . As such apps are not authorative sources of truth and only used for secondary purposes, a small positive error rate in the output can be acceptable. Stand-alon
AI-assisted fragment prioritization was connected to NMR and X-ray evidence rather than evaluated only in silico.
Our approach to bioresilience The global biosecurity landscape is rapidly evolving. Shifting natural ecosystems, global travel and the potential misuse of AI require greater vigilance — yet AI is also a critical tool for our response. We need frontier AI models, and the scientific advances they will enable, to respond to these challenges and help make society more resilient to events like future outbreaks. Today, Goo
Fluorescent proteins and small-molecule dyes offer complementary advantages for biological imaging: Proteins are amenable to genetic tagging, whereas dyes provide superior brightness and photostability. To combine these strengths, we used de novo protein design to generate small, nanomolar-affinity, high-selectivity binders (NovoTags) for three cell-permeable dyes spanning the visible spectrum. We show that the NovoT
Conjecture Machines: AI agents and the new validation bottleneck in science Over the past year, agents have transformed what it means to code. Scientific research may be next — from proposing novel hypotheses, to designing experiments, to discovering algorithms that improve on the best that humans have designed. This raises urgent questions for policymakers and science funders, the biggest being how to validate the c
SemiEpi connects a physics model, in-situ sensing and immediate feedback to control molecular-beam epitaxy.
Terry Tao demonstrates how coding agents facilitate converting code-generated figures into interactive applets, using Figure 1 from his recent paper on the Gilbreath conjecture with Chase and Hunter. The figure plots exact and Monte Carlo simulated values of a sequence called the "Gilbreath expectation sequence," which arises from a Gilbreath array: an inverted pyramid with top entries as independent exponential rand
I am finding the newly revealed capability to code old applet ideas into reality to be very tempting to sink more time into, though I am certainly encountering the common “vibe coding” experience that the process can produce something that superficially resembles a finished product well before a satisfactory level of testing and review has been completed; indeed, it is the review process which is now the most time-co
How Rust, Lean, Aeneas, and AI agents are helping scale formal verification for production cryptographic algorithms At a glance SymCrypt develops new verified cryptography using Rust, Aeneas, and Lean to provide higher security assurance. We prove that their code safely and correctly implements standard algorithms, notably for post-quantum cryptography. We are releasing verified code, specs, properties, and proofs in
In a 25-dimensional, 528-run closed-loop experiment, expert chemists and LLMs were compared directly, producing a useful mixed result for AI-scientist claims.
I have been interested in machine-assisted ways to do and teach mathematics from as far back as 1999, when I started coding several applets in Java 1.0 , both for my complex analysis and linear algebra courses, as well as to visualize various mathematical objects I was interested in (such as honeycombs or Besicovitch sets). This was moderately successful; but the applets were time-consuming to program. Eventually, th
Herein, we combine Bayesian Optimization and mechanochemistry to develop a rapid, solvent-free protocol for the Johnson–Corey–Chaykovsky (JCC) cyclopropanation and epoxidation. By prioritizing efficiency and sustainability, Machine Learning quickly identified a new set of reaction conditions for this venerable reaction that diverge from established paradigms (explosive NaH/DMSO system), revealing that these transform
Biomni integrates tool retrieval, code execution and experimental protocols across biomedical tasks in a single agent environment.
Microsoft Research AI for Science introduces Flint, a visualization intermediate language for AI-driven chart creation. Flint enables AI agents to reliably generate expressive, visually polished charts from simple, human-editable specifications. It leverages semantic data types to guide the compiler in choosing appropriate scales, baselines, formatting, and color schemes, and automatically manages layout to adapt to
Abstract Self-driving lab systems (aka, autonomous experimentation) accelerate research - letting scientists learn faster, spend less resources, and fail smarter in well defined, narrow studies. The next-generation materials lab combines self-driving systems to tackle broader challenges - orchestrating complex research campaigns while optimizing lab resources. We propose that agent-based and agentic artificial intell
SkillOpt treats an agent's skill file as a trainable parameter outside a frozen target model, converting skill writing from one-shot prompting into a controlled optimization process. Across six benchmarks, seven target models, and three execution modes, SkillOpt achieves best or tied-best performance in all 52 evaluation cells, improving performance without updating model weights. The method keeps skills compact and
XtalPi has entered a strategic partnership valued at more than $400m with an international biopharmaceutical company to develop oral small-molecule therapeutics targeting a G protein-coupled receptor (GPCR). The collaboration follows a successful pilot phase, where XtalPi’s integrated quantum physics and AI algorithms achieved improved hit rates, highlighting the platform’s capability to address the complex metabolic
Science Jun 8, 2026 Written by Laura Luebbert. Based on research by Ferdous Nasri, Sarah Gurev, Patrick Varilly, Krithik Ramesh, Nuala A. O’Leary, Jonah Cool, Bernhard Y. Renard, Pardis Sabeti, and Laura Luebbert. In this post, Laura Luebbert argues that we need to make biological data infrastructure more agent-friendly. As a case study, she and her team tasked scientific research agents (Claude, Biomni Open Source (
Last week, OpenAI announced that an internal AI model had disproved the Erdős unit distance conjecture, a famous problem in discrete geometry that had stumped human mathematicians for the last 80 years. OpenAI gave several mathematicians early access to the result and published their reactions. Tim Gowers — who won the Fields Medal, the most prestigious prize in mathematics — wrote that “there is no doubt that the so
A Stanford-led research team that includes Fei-Fei Li published a benchmark on May 18, 2026, that documents a specific, measurable failure in every frontier multimodal AI model tested: when forced to actively move through a 3D environment rather than interpret a pre-composed image, state-of-the-art systems consistently choose the wrong actions, skip viewpoints that would correct their mistakes, and commit to wrong an
I’m organizing a workshop in London on July 6th to 10th (2026) whose goal is to work on my EPSRC-funded project formalizing Fermat’s Last theorem in Lean. The initial aim of the project was to reduce FLT to theorems known in the 1980s, i.e. “formalize the Wiles/Taylor–Wiles papers” (although actually we are formalizing a more modern proof whose ideas go back to a strategy proposed by Khare). I have just finished givi
Demis Hassabis, CEO of Google DeepMind and 2024 Nobel laureate in Chemistry, has spent the first months of 2026 articulating a specific and measurable thesis: AlphaFold, the protein-structure model that now serves more than three million researchers across 190 countries, is the first rung in an AI-powered scientific revolution, not the last. His clearest statement of this position came in a Semafor interview publishe
Google DeepMind presents an agentic system designed to collaborate with research mathematicians on open-ended problems.
Science Apr 29, 2026 In this post, Brianna, a researcher on the discovery team, shares results from a recent bioinformatics benchmarking effort. Almost as soon as large language models could hold a conversation, people started asking how they’d stack up against human experts. Could models pass the bar exam? Could they answer medical licensing questions, or solve Olympiad math problems? Such benchmarks—self-contained
Jùn Zhang and Carina Hong discuss translating mathematics into Lean, formal proof, mathematical intuition and building an AI-for-math company.
Computational X-ray absorption near-edge structure (XANES) is widely used to interpret local coordination environments, oxidation states, and electronic structure, but large computational campaigns are often limited by workflow complexity. We present ChemGraph-XANES, a large language model (LLM)-based agentic framework that combines documentation-grounded parameter retrieval via retrieval-augmented generation (RAG),
The Research and Applied AI Summit (RAAIS) is a community for entrepreneurs and researchers who accelerate the science and applications of AI technology. In the run up to our 10th annual event on June 12th 2026 in London, we’re running a series of speaker profiles to shed more light on what you can expect to learn on the day! Vivek Natarajan is a Research Lead at Google DeepMind leading research at the intersection o
Image credit: Elianne Dipp / Pexels Over the last decade, work on animal communication has reached a convergence as the growing availability of large-scale behavioral and bioacoustic data merges with relatively recent advances in machine learning. As a result, we’re starting to see the study of animal communication shift from narrowly defined, species-specific, and hypothesis-limited analyses toward data-driven disco
Science Mar 23, 2026 Can AI do theoretical physics? In this guest post, professor of physics Matthew Schwartz decided to find out by supervising Claude through a real research calculation, start to finish, without ever touching a file himself. His account of what happened is below. - I guided Claude Opus 4.5 through a real theoretical physics calculation, encapsulating the complexity of code and computations behind t
Science Mar 23, 2026 In this post, Siddharth Mishra-Sharma, a researcher on the Discovery team, explains how to apply multi-day agentic coding workflows—test oracles, persistent memory, and orchestration patterns—to scientific computing tasks even outside of one’s domain. Most scientists currently using AI agents work in a conversational loop, managing each step of the process on a tight leash. As models have become
James Zou | Associate Professor of Biomedical Data Science, Stanford University, California, USA Citation: EMJ. 2026;11[1]:28-31. https://doi.org/10.33590/emj/OR9BWSEN When did you first become interested in applying AI to clinical research? I’ve been working at the intersection of AI and biomedicine for about 15 years now, since my PhD. My initial work was more focused on using AI to enable biomedical discoveries. W
OpenAI’s GPT can summarize research papers and make predictions—but can it do science? Can it generate hypotheses, design experiments, interpret results and iterate? Last summer researchers at OpenAI and Ginkgo Bioworks, a company that designs and installs autonomous, robot-run labs, decided to find out. Though artificial intelligence systems have posted high scores in math, physics and computer science, biology is h
Demis Hassabis traces how AlphaGo's search and learning ideas fed into AlphaFold, AlphaProof, AlphaEvolve and the broader scientific-AI program.
Optimizing molecules to achieve desired properties is a central bottleneck across the chemical sciences, particularly in the pharmaceutical industry, where it underlies the discovery of new drugs. Since molecular property evaluation often relies on costly and rate-limited oracles, such as experimental assays, molecular optimization must be highly sample-efficient. To address this, we introduce SEISMO, an LLM agent fo
Accelerating Mathematical and Scientific Discovery with Gemini Deep Think Under direction from expert mathematicians and scientists, Gemini Deep Think is solving professional research problems across mathematics, physics, and computer science In the summer of 2025, an advanced version of Gemini Deep Think achieved Gold-medal standard at the International Mathematics Olympiad (IMO) and later, an updated version, obtai
Google DeepMind’s AlphaEvolve AI is accelerating mathematical research at an unprecedented scale, according to a new paper published this week by collaborators including renowned mathematician Terence Tao. The research shows how the AI agent tackled 67 challenging problems, rediscovering top solutions and finding novel constructions for several long-standing challenges. The work demonstrates a powerful new method for
So it’s an interesting time for computers-doing-mathematics. A couple of interesting things happened in the last few days, which have inspired me to write about the question more broadly. First there is the question on whether computers will ever prove interesting research-level mathematical theorems autonomously at all, which is really hard to answer right now, although clearly a lot of people are betting on “yes”.
Sam Rodriques, an alum of the Synthetic Neurobiology research group, has been recognized in the Innovators category of the TIME100 AI 2025 list. Dr. Rodriques co-founded the nonprofit FutureHouse in 2023, with the aim of giving scientists new tools to break through bottlenecks and improve their workflows. To that end, FutureHouse is developing AI agents that can formulate hypotheses, conduct experiments, and analyze
FutureHouse's head of science explains its agent stack, a drug-repurposing case, and why scientific publishing and evaluation must be redesigned for machine readers.
A theoretical physicist by training, Sam Rodriques recognized the possible uses of AI for scientific discovery. In 2023 he co-founded the nonprofit FutureHouse, with a goal to build AI scientists that can formulate hypotheses, conduct experiments, and analyze data by themselves. While AI science is an increasingly crowded field, FutureHouse’s audacious vision and scientific bonafides have earned it the backing of Eri