Your AI research companion

1. cs.LGOctopus: One Brain, Many Branches for Latent-KV Multi-Agent Reinforcement LearningTianming Sha

   The authors built a system in which one language model can run several coordinated research threads while sharing its memory instead of repeatedly copying it. It achieved higher scores on two research tasks than the compared approaches while using less …

2. cs.LGSteelAgent: LLM-Orchestrated, Physics-Informed Steel Property Prediction with Audited Generalization and Cali…Aleksandr Volkov, Roman Sultimov, Mikhail Kuzin, Yury Maximov

   The authors examined how well steel-strength predictions hold up when models face unfamiliar groups of steels rather than randomly chosen test examples. Performance and uncertainty estimates worsened sharply in these harder tests, showing why reported predictions should specify what kinds …

3. cs.AIWhen Does Test-Time Search Help World Action Models? An Empirical Audit of Candidate Selection and Simulator …Jun Liu, Dajie Sun, Zhenglun Kong, Zhenjiang Fan, Peiyan Dong, Fan Yang et al.

   The authors tested whether spending extra computation to choose among action plans really improves robot control, saves time, and produces trustworthy evaluations. In the tested systems, selection gains were weak, actual speedups were limited, and a hidden rendering error affected …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

New preprints explore robots learning from simulation, AI-assisted mathematical discovery, and why accuracy depends on what—and where—we choose to measure.

6 papers put into perspective

1. The coastline looks different outside the training set
2. When more accurate arithmetic breaks a model
3. AI text spreads unevenly across the web
4. Cleaner movement data can change the movement being measured

1. astro-ph.GAAGN Archaeology: A Systematic Search for Fading AGN in MaNGAQuinn O. Casey, Ryan C. Hickox, Jonathan H. Cohn, Nikko J. Cleri, Julia M. Come…

   The authors use spatially detailed galaxy observations to find 34 galaxies whose central black holes appear less active now than their surroundings suggest they were in the past. These fading systems offer clues to how long black holes stay active …

2. astro-ph.HEGeneral-relativistic neutrino-radiation hydrodynamics of crust ejection in magnetar giant flaresTetyana Pitik, David Radice, Jacob Fields, Yi Qiu

   The authors simulated how powerful magnetar flares could blast material from a neutron star’s crust and calculated what elements that material might produce. Their results suggest that even strong flares eject little heavy-element material, and that realistic flares may contribute …

3. astro-ph.GAThe obscured nucleus of NGC3256 S. I. Anisotropic AGN near-IR radiation field traced by CO v=1-0 with JWST/NI…Miguel Pereira-Santaella, Eduardo González-Alfonso, Ismael García-Bernete, Andr…

   Using JWST observations and radiation models, the authors studied how infrared light escapes from the obscured center of the nearby galaxy NGC 3256. They found that the light is directed into a narrow path and that the central black hole …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

This week’s preprints uncover an overlooked outer Solar System object, revisit a stubborn chemical puzzle and use light from stellar remnants to reconstruct what cannot be seen directly.

6 papers put into perspective

1. A bright world in a neglected patch of sky
2. A chemical discrepancy becomes a temperature problem
3. Measuring a neutron star across the spectrum
4. A supernova with a delayed return

1. quant-phSingle-shot ground-state preparation via strongly coupled bathsLiang-Hong Mo, Jerome Lloyd, Yuxuan Zhang, Dmitry A. Abanin, Biao Lian

   The authors develop a way to cool quantum many-body systems quickly by coupling them strongly to a surrounding bath, rather than relying on slow, weak interactions. Their protocols can prepare ground or thermal states in a single cycle and are …

2. quant-phThreshold-Free Symmetry-Enriched Topological Quantum MemoryAbhijat Sarma, Ruochen Ma, Cenke Xu

   The authors propose a way to protect stored quantum information in a toric-code memory even when errors are very frequent. Their construction uses additional symmetry-related degrees of freedom and a decoder that can correct errors across the full range of …

3. quant-phNonlocal magic of fermions is bounded by entanglement fluctuations and is logarithmic for free fermionsRaghav G. Jha, Dengyao Luo, Alexander F. Kemper, Bojko N. Bakalov

   The paper investigates a measure of quantum resource in free-fermion states, which are often considered easy to simulate on classical computers. It finds that most of this resource can be removed by suitable local operations, and that the part that …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

New preprints explore how quantum behavior is lost, preserved and engineered—and how a seemingly reversible change can undermine quantum communication.

6 papers put into perspective

1. Watching an electron lose its quantum character
2. For silicon qubits, quieter can also mean harder to drive
3. A small quantum processor hosts a fractional quantum state
4. Repeated measurement slows a spin system’s forgetting

1. cond-mat.mtrl-sciSudden disappearance of a bright exciton in solid heliumFatema Mohamed, Lucia Reining, Matteo Gatti

   The study examines how light interacts with bound electron-hole pairs in solid helium when the probe direction changes. It finds that a bright spectral feature can disappear completely in certain directions because of interactions between the electron and hole, showing …

2. physics.opticsReliability of Fisher-information bounds for dielectric metalenses: a site-resolved RCWA auditJ. Sumaya-Martinez, O. Olmos-Lopez

   The authors checked whether a computer-based estimate of how precisely a metalens can locate an object is reliable. Their calculations closely reproduced the published estimate, while also showing that assumptions about the lens and detector matter and that the result …

3. cond-mat.softPFAS-induced modulation of vesicle diffusiophoresisSeungsu Han, Gbassey S. A. Oteme, Prathima Nalam, Sangwoo Shin

   The researchers tested how PFAS chemicals affect the charge and movement of tiny fat-based bubbles that model cell membranes. PFAS made the bubbles move more strongly in chemical gradients and caused some membrane material to be lost, offering clues about …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

This week’s selected preprints use interference, ultrafast light, electron correlations and living swimmers to reveal how microscopic behavior becomes measurable collective order.

6 papers put into perspective

1. Distant clocks become a probe of gravity
2. A superconducting state reveals its internal sign pattern
3. Light opens a topological gap in graphene
4. Clockwise swimmers build a counterclockwise fluid

1. hep-thConformal Bootstrap UnblockedLi-Yuan Chiang, David Poland

   The authors introduce a way to test theories of how correlations behave in systems with conformal symmetry without using the usual toolkit based on operator expansions. They demonstrate that it can reproduce familiar patterns in free-field and Ising models and …

2. hep-thAsymptotic safety is not genericFrancesco Del Porro, Francesco Ferrarin, Alessia Platania

   The authors tested whether a proposed way to complete quantum gravity at very high energies also appears in other theories with some similar mathematical features. In the models they studied, they found no evidence for the required fixed point, suggesting …

3. hep-thFermion$\text{-}$Monopole Scattering: A Unified PerspectiveHao Geng, Zekai Wang

   The paper brings together different explanations of what happens when a massless fermion scatters off a magnetic monopole. It shows how the monopole's internal motion, the form of the outgoing fermion, and unusual fermion correlations fit into one consistent account …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

New preprints use neutrinos, nuclear decay, colliding nuclei, numerical QCD and gravitational waves to test how well our descriptions of nature survive demanding measurements.

6 papers put into perspective

1. A clearer signal from reactor neutrinos
2. Small nuclear collisions show a collective pattern
3. An exotic nucleus tests how particles decay together
4. Computing the thermodynamics of dense quark matter

1. q-bio.NCCategory differences in similarity between fixation density maps and psychological potential fields for sketc…Naoyuki Awano, Yuki Hayashi

   The study compared where people look while viewing sketches with a map intended to represent aspects of visual understanding, across meaningful objects and meaningless shapes. Adjusting the comparison for baseline differences revealed clear category differences, while rotating the sketches did …

2. q-bio.NCApproxBrain: Certified Sufficiency Frontiers in Connectome-Constrained Neural SystemsAli Mahdavi, Azadeh Zamanifar, Amirfarhad Farhadi, Omid Kashefi

   The authors developed a way to describe which groups of neurons are enough for a task, while accounting for acceptable performance and variation in model settings. Tests in a walking simulator found that the required neurons depend strongly on the …

3. q-bio.NCA minimal dynamical model of human-AI cognitive coevolutionPetar Jolakoski, Dimitar Trajanov, Qiyu Sun, Zhen Wu, Ljupco Kocarev

   The paper builds a mathematical model of how AI use might affect human reasoning over time, distinguishing the ability to solve new problems from knowledge built through experience. The model shows possible conditions under which AI could either support cognition …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

A simulated fly, a stress test for seizure prediction and studies of developing brains and inherited risk reveal why biological explanations depend on what surrounds the system being studied.

5 papers put into perspective

1. A fly connectome takes its first steps
2. The seizure predictor depends on the test
3. Seeing helps blood vessels mature
4. Birth changes the developmental clock

1. bioRxiv: Cell BiologyAltered Steroidogenesis in Response to Impaired Mitochondrial Protein HomeostasisM. B. Slavin, L. Lee-Glover, T. Zhao, M. Uddin, A. Joselin, R. Wong et al.

   The study tested how different levels of stress on mitochondria affect cortisol production. Mild stress raised cortisol, while severe stress lowered it, suggesting that mitochondrial stress responses may influence hormone levels in conditions such as Parkinson’s disease.

2. bioRxiv: Cell BiologyKDM6A is enzymatically active at the Golgi apparatus where it is necessary for maintenance of Golgi apparatus…C. S. Worth, J. M. Conforti, A. M. Nambiar, J. Malone, R. Budhathoki, A. M. Zie…

   The study found that KDM6A, a protein usually discussed for its role in controlling genes, can also remain active in the Golgi apparatus, a cell structure that processes and moves materials. Reducing KDM6A changed Golgi shape, suggesting that the protein …

3. bioRxiv: Cell BiologyThe Application of 3D Cryo-Volume Electron Microscopy to Visualize and Quantify the Photosynthetic Filamentou…A. R. Angermeier, L. G. Rotkina, C. Lebow, W. Chiu, T. M. Burch-Smith, K. J. Cz…

   The authors tested a way to image whole cyanobacteria in three dimensions while keeping their structures frozen and avoiding chemical treatment. The resulting images revealed internal cell features and allowed measurements of structures involved in photosynthesis and storage.

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

This week’s preprints trace how nerves, immune cells, fungi and membrane proteins shape disease—and how a diagnosis can remain absent even when the laboratory evidence is present.

6 papers put into perspective

1. A nerve reflex opens the lung’s vascular barrier
2. An adjustable doorway for folded proteins
3. Engineered immune cells need reinforcements
4. A hemorrhagic fever virus depends on host entry receptors

1. physics.ao-phA Deep-Learning Framework for the Generalized Emulation of Atmospheric Radiative Transfer at Microwave Freque…Spencer R. Jones, Simon Pfreundschuh, Christian D. Kummerow

   The authors trained neural networks to imitate calculations of how the atmosphere and Earth’s surface affect microwave signals. The emulators produced results more than a thousand times faster than the reference calculations while keeping errors very small.

2. astro-ph.EPResults from a Stellar Occultation by (623) ChimaeraPhillip Oakey, Keya Garg, Andrey Moore, Altony Foote, Michael F. Skrutskie, Ann…

   The team observed the asteroid Chimaera pass in front of a star, using measurements from telescopes spread across the United States to map its outline. They estimated its size and reflectivity, and found that computer models can predict how it …

3. astro-ph.EPAn analytical framework for atmospheric tides on rocky planets. II. Semidiurnal thermotidal torquePierre Auclair-Desrotour, Jacques Laskar, Mohammad Farhat, Gwenaël Boué

   The authors develop an analytical way to predict how a rocky planet’s atmosphere responds to heating from its star and how that response affects the planet’s spin. Their predictions match earlier computer simulations and help explain how atmospheric waves can …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

New preprints examine what forest treatments protect, how parasites silence collective defenses, and what clocks, comet dust and weather models can reveal about a changing world.

5 papers put into perspective

1. Saving trees without restoring a fire-adapted forest
2. A parasitoid cuts the alarm line
3. Interstellar dust with little sign of heating
4. Clocks follow the changing pull of the tides

1. math.OCPractical Semialgebraic Optimization via Sparse Sample-Based Sum-of-Squares ProgrammingJason Nezvadovitz, David M. Rosen

   This paper develops a way to make certified optimization methods work on larger problems by using carefully chosen samples and exploiting local structure. It demonstrates the approach on applications including 3D matching, which it solves quickly on an ordinary laptop.

2. math.PRA second eigenvalue lower bound for the directed configuration modelCharles Bordenave, Yizhe Zhu

   The authors prove a lower limit on how slowly a random directed network’s random walk can mix, with high probability. Their proof uses patterns in repeated matrix operations rather than requiring a full description of the network’s eigenvalue distribution.

3. math.PRNecessary and sufficient conditions for the martingale property of stochastic exponentials driven by multi-di…Erhan Bayraktar, Hao Xing

   This paper gives conditions for deciding when a commonly used exponential expression built from a multidimensional random process has a valid probability-like behavior over time. It applies the criteria to financial models and also connects them to uniqueness of pricing …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

New preprints revisit black-hole thermodynamics, geometric efficiency, numerical algorithms and the limits of what simple models can represent.

6 papers put into perspective

1. An extremal black hole, reached through vacuum
2. The geometry of enclosing space
3. A speed limit for finding roots
4. A community can fail despite reassuring interaction signs

1. cs.SEHow Python Audio Libraries Change Across Releases: Module Structure and DocumentationMateusz Modrzejewski, Michał Rokita, Łukasz Neumann

   The authors track how Python audio libraries change from release to release and whether release notes describe those changes. They find that module removals are rarely mentioned in the notes, giving developers evidence that upgrade documentation may miss important structural …

2. cs.SEBeyond the Agent Harness: Boundary Criteria for Software-Factory Orchestration - A Single-Operator Case StudyLei Hua

   The paper proposes rules for deciding whether a software-development mechanism or failure belongs to an individual coding agent, a system coordinating many agents, or another layer. Applying the rules showed strong coder agreement on classifications, but also frequent ambiguity when …

3. cs.SEA Multi-Modal Robotic Framework for Physical GUI Interaction TestingDavi Freitas

   This work proposes a robot-based system that tests smartphone apps by looking at the screen and physically tapping, swiping, and typing, rather than using hidden software controls. Early tests show it can recognize common screen elements, which could make app …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

Among this week’s computing preprints, progress comes from reorganizing calculations, fitting speech recognition onto a microcontroller, and testing whether elaborate systems actually earn their complexity.

6 papers put into perspective

1. A famous coloring problem gets a faster solution
2. Counting a disordered system up to a critical boundary
3. Speech recognition without a neural accelerator
4. A newer GPU is not automatically a better purchase

1. econ.EMUnobserved Consideration and Consumer WelfareDebopam Bhattacharya, Ryan J. Martin

   The paper asks whether researchers can estimate how price changes affect consumer welfare when they do not observe which products each person considers. It shows welfare can still be identified under stable limits on consideration, and gives bounds when higher …

2. econ.EMWillingness-to-pay space and preference space: a unified treatment of equivalence, heterogeneity and interpre…Stephane Hess

   This note explains when two common ways of expressing consumer choice models are simply different versions of the same model, and when they are not. It clarifies how to interpret willingness-to-pay estimates, especially when preferences vary across people or respond …

3. q-fin.STNeural Realized Measures: Forecast-Supervised Intraday Aggregation for Volatility ForecastingBrian Sing Fan Chan, Jennifer So Kuen Chan

   The authors develop a method that learns how to combine intraday market data for volatility forecasts instead of using a fixed averaging rule. In tests, it improves forecasts across simulated data and six cryptocurrencies, suggesting that the way daily measures …

Every weekday, the new papers in your categories, each with a short summary. See a whole digest →

New preprints examine maternity leave, agriculture’s origins, policy evaluation, high-speed trading and bank capital—and show why the route from data to conclusions matters.

5 papers put into perspective

1. Paid maternity leave and the limits of a policy cutoff
2. Why farming began—and why it spread
3. A feared statistical problem meets a large empirical check
4. What trading speed buys in China