Wednesday, October 7
2,322 new papers appeared. We read every one. These 20 are worth your time.
Today's edition
Picked for broad interest, a real step forward and honest claims, with no field allowed to take over.
Global and regional reference curves for pain across the lifespan in 6.1 million individuals in 118 countries
In this study, we harmonized individual-level self-reported pain data from 6,125,459 participants across 902 population-based data sources in 118 countries to establish global reference trajectories of pain.
AI summary
Pain rates varied from about 2% for facial pain to 40% for back pain, with women reporting higher pain at all 11 body locations. Unlike previous estimates, people in the poorest countries had much higher pain in old age than those in the richest countries, with a 30.6 percentage point difference. Globally, smoking, obesity and low income explained 18.3% of pain burden, but this ranged from 12.6% in sub-Saharan Africa to 27.1% in eastern Europe. The authors made these patterns available in a free online tool for comparing other datasets against global pain norms.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalClaims a big stepReal-world useLarge and persistent social risks to forest carbon projects
Here we analyze 244 forest-based carbon projects to assess the vulnerability of the voluntary carbon market to social risks and find that projects systematically underestimate their risk of failure.
AI summary
Carbon credits are popular for offsetting emissions, but their climate claims face growing doubts. The authors analyzed 244 forest carbon projects and found they systematically underestimate their failure risk. Even optimistically, projects are about twice as likely to fail within 25 years as they claim over 100 years. Credits from failed projects could exceed registry reserves within 1-2 decades. Stronger mechanisms and reserves may be needed to ensure credits deliver promised climate benefits.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintBroad interestReal-world useEasy to readChiral discrimination of amino acids in meteorite and desert soil extracts via single-molecule nanogap conductance
Here we show that quantum tunneling currents electrically distinguish individual L- and D-amino acids and quantify their ratios at the single-molecule level without optical measurements or chiral recognition molecules.
AI summary
Life on Earth uses left-handed amino acids, while non-living processes make equal amounts of both mirror-image forms. The authors used quantum tunneling currents to tell apart individual left- and right-handed amino acids electrically. They identified single enantiomers with over 80% accuracy and analyzed extracts from the Murchison meteorite and Atacama Desert. This electrical method provides a foundation for studying complex samples in astrobiological research.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalDeep learning enables large-scale inverse design of free-form metasurfaces
Here we introduce a physics-tailored deep-learning surrogate model that exploits the locality and symmetries of electromagnetic interactions to generalize from small-scale simulations to metasurfaces orders of magnitude larger.
AI summary
Metasurfaces are ultrathin optical elements that control light through tiny structures smaller than light's wavelength. Designing large metasurfaces was too slow because each design needed repeated electromagnetic simulations. The authors created a deep-learning model that learns from small simulations and applies to much larger metasurfaces. They demonstrated this by building free-form metagratings and a millimeter-scale holographic metasurface with 1.6 billion degrees of freedom.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepReal-world useSix objects reported at 480 mph in a declassified military infrared video show brightness flicker at bird wingbeat rates
The six objects are best explained as a small flock of birds flying nearer to the aircraft than the crew assumed.
AI summary
A 2026 military video shows six objects moving at 480 mph. The objects flicker steadily 7.1-7.9 times per second, each at its own rate. Tests show the flicker is not caused by video compression or the sensor. The authors conclude the objects are likely birds flying closer than assumed.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintBroad interestEasy to readA Primer on Ethics in Mathematics
It develops much of the foundational theory of ethics in mathematics: what can go wrong, why it goes wrong, and how to address it.
AI summary
This paper introduces a new course on ethics in mathematics. The course covers ethical issues in mathematical work and mathematically-powered technologies. It uses case studies of harmful behavior and interactive activities for learning. The approach is practical rather than deeply theoretical. It is relevant to computer science, physics, engineering, and other numerate fields.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintEasy to readLarge scientific teams are more, not less, disruptive
For scientific publications, we conclude that larger teams are more, not less, disruptive.
AI summary
A 2019 study claimed small teams produce more disruptive science than large teams. The authors show this finding was mainly a measurement problem, not a real difference between teams. Larger teams simply cite more references, which mechanically lowers the standard disruption score. When they fixed this bias, larger teams proved more disruptive across four fields and every decade since 1970. Their new method also better identified actual breakthroughs, including Nobel Prize-winning work.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintBroad interestSecure Speculative Decoding for Large Language Models
In this work, we bridge this gap by providing the first systematic study of the security implications of speculative decoding.
AI summary
Speculative decoding speeds up large language models by using a smaller draft model to create candidate tokens that the main model then checks. The authors found that faster decoding methods sharply increase security risks from jailbreak and prompt injection attacks compared to small drops in usefulness. They created SecureSD, which applies stricter checks to early tokens from the draft model based on their theory. Tests showed SecureSD improves security while keeping speed and usefulness compared to other speculative decoding methods.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepReal-world useIllusory Pattern Perception Drives Spurious Inference in Large Language Models
We find that LLMs frequently exhibit stronger illusory pattern perception than humans.
AI summary
This study tests whether large language models see imaginary patterns in random data, like humans do. The authors adapted classic psychology tests to compare model and human behavior directly. They found that models show stronger illusory pattern perception than humans. Models often link frequent positive traits to majority groups and build causal stories from unclear events. Using a new analysis tool, the authors linked this behavior to how models perceive frequency and process information.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintCode availableA broadly conserved phytoplasma α/β-hydrolase effector, SGP-7, is a catalytically active esterase that hydrolyses strigolactone analogues and alters shoot architecture
Together, these results define the biochemical function of a conserved phytoplasma effector family and identify strigolactone-like molecules as direct enzymatic targets of a bacterial plant pathogen.
AI summary
Phytoplasmas are bacteria that live in plant sap and change how plants grow by releasing effector proteins. The authors studied SGP-7, a protein found in at least seven groups of these bacteria. When the researchers put SGP-7 into plants, the plants grew shorter and developed more side shoots.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepProtein-level mapping reveals novel Immune State Types (ISTs) associated with Chlamydia trachomatis infection, inflammatory remodeling, and scarring in trachoma-hyperendemic Amhara Ethiopia
Conclusions: This study establishes the first IST framework for trachoma, revealing biologically distinct immune states with different potential clinical implications.
AI summary
Researchers studied 667 people in Ethiopia one year after mass antibiotic treatment for trachoma. They measured 41 immune proteins in eye tissue and found four distinct Immune State Types. One state showed focused immune response to Chlamydia infection, while another showed broad inflammation. Two scarring states differed: one had active chemokine signals, the other had quiet immune activity. These immune states may guide treatment choices, from antibiotics to possible immunotherapies or antifibrotic drugs.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepReal-world useZeppelin: Client-Side BFV Encryption and Decryption for Helium-Powered Microcontrollers
We present Zeppelin, the first HE library to use an embedded vector instruction set, the first to support both encryption and decryption on an embedded device, and the first BFV implementation on MCU-class hardware.
AI summary
Zeppelin is a new encryption library for small Internet-connected devices. It uses BFV, a method that works with whole numbers and recovers exact values. It speeds up the main encryption step using vector instructions in ARM's Helium hardware. On one small chip, it encrypts 4096 values in 4.76 ms and decrypts in 5.38 ms using under 500 KB of memory.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepReal-world useLayered spin-crossover metal-organic frameworks for light-induced control of two-dimensional quantum materials
These results establish layered stimuli-responsive SCO MOFs as active molecular control elements for 2D quantum materials, providing a route to strain-mediated control of electronic, magnetic, and superconducting functionalities and extending MOF-based architectures beyond traditional porous-matter applications towards multifunctional electronics and spintronics.
AI summary
The researchers integrated a layered spin-crossover metal-organic framework into electrical nanodevices to control two-dimensional quantum materials. In graphene, light selectively switched the conductivity. In magnetic CrSBr, the material triggered magnetic hysteresis and enabled tunable non-volatile zero-field memory. In superconducting NbSe2, light modulated the superconducting critical current and transition temperature.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Preprint with a published versionFirst Direct Identification of a Multi-Star Microlens System Hosting a Planet
This is the first direct detection of flux coming from a two-star microlensing system which hosts a planet.
AI summary
Scientists studied images of a microlensing event called OGLE-2006-BLG-284 taken more than 15 years after it happened. They detected the lens system and measured how it moved across the sky. The system contains two stars with masses of 0.43 and 0.12 solar masses, and a planet with a mass of 165 Earth masses. The authors could not determine whether the planet orbits one star or both stars. This is the first direct detection of light from a two-star microlensing system that hosts a planet.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepLearning in Dreams, Winning in Reality: A Continuous Dyna Loop for a Ten-Hero MOBA
Run as a continuous asynchronous Dyna loop, the policy wins 70.2% of real games as radiant (421 of 600; 95% CI 66.4-73.7) on seeds never used for any decision, up from 0% for dream training alone and 33.7% before the loop.
AI summary
The authors built a world model for a ten-hero MOBA game with 206 units and games up to 6,000 ticks. They trained a policy only inside imagined episodes, then tested it in the real game without using real-game gradients. The policy won 70.2% of real games as radiant, up from 0% with dream training alone and 33.7% before the loop. Model exploitation was invisible from inside the dream: unanchored runs collapsed without any in-dream metric showing it. The policy inherited the model's limits: it won through map-wide pressure, not coordinated fighting.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepCode availableLeanPlan: Optimal Planning with LLM-Generated Heuristics and Admissibility Proofs
We introduce LeanPlan, the first planning system that finds optimal plans with LLM-generated heuristics whose admissibility is machine-checked.
AI summary
Large language models can create heuristic functions that guide search for planning tasks, but these heuristics may not be admissible and can produce suboptimal plans. LeanPlan is the first system that finds optimal plans using LLM-generated heuristics with machine-checked admissibility proofs. The system uses an agentic loop with planner feedback to iteratively improve a reusable domain-specific heuristic, its proof, and required domain assumptions. LeanPlan implements the heuristic, proof, and planner with machine-checked grounding and search in Lean 4. Evaluated on thirteen domains with test tasks up to 57 times larger than training tasks, LeanPlan with GPT-5.6 Sol successfully generated heuristics and proofs for all domains. With these heuristics, LeanPlan usually expands fewer states than the state-of-the-art Scorpion planner and solves more tasks overall.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepFormation of a swelling gel underlies a morphological transition in Bacillus subtilis biofilms
Here, we show that Bacillus subtilis biofilms use polymer composition to tune their physical properties and drive gel formation.
AI summary
Bacillus subtilis bacteria tune their biofilm properties by adjusting which polymers they produce. One polymer absorbs water and swells the colony, while another acts as cross-linkers to create a gel. Colonies producing both polymers develop macroscopic wrinkles. A thin-film model predicts this wrinkling from swelling-induced strain combined with elasticity. The model matches the observed shape of the morphological phase diagram.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalA Technosignature Survey Of The Anti-Solar Point: The SETI Institute's Largest-Ever Campaign With The Allen Telescope Array
Although SETI literature has long proposed searching for beacons at the ASP, this campaign provides the first observational coverage.
AI summary
This point is a 'Schelling point'—a location aliens might also choose because they could see Earth crossing its star from there. The survey has covered about 12% of its target area so far, with 67% of sessions analyzed. Researchers found 25 narrowband drifting signals located at the Anti-Solar Point and will re-observe those 10 spots. This is the first time anyone has actually observed this proposed location for alien beacons.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintA map of RBP-miRNA regulatory connections for deciphering the mechanisms of transcriptome remodeling in cancer
Here, by modelling the regulon activities of well-characterized miRNA families in RNA-seq data of a comprehensive collection of RBP inhibition experiments, human cell lines and tissues, we report a map of regulatory connections between 34 RBPs and 28 miRNAs.
AI summary
The authors mapped connections between 34 RNA-binding proteins and 28 micro-RNAs using data from experiments where RBPs were blocked. They found that RBP-driven regulation of micro-RNA activity is common in tumors and adds to differences in micro-RNA activity patterns. They identified RBPs that control tumor-suppressor and cancer-promoting micro-RNAs, including miR-29-3p, miR-17-5p and miR-19-3p. The protein DDX55 reduces the activity of the tumor-suppressing miR-29-3p family across many cancer types. Blocking DDX55 restored miR-29-3p's gene-silencing ability and lowered cancer cell survival.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepElectrical stimulation of the human pulvinar generates visual percepts
These findings suggest that pulvinar stimulation can generate visual percepts and modulate distributed cortical networks, supporting further investigation of the pulvinar as a target for vision restoration.
AI summary
Researchers tested whether electrically stimulating the pulvinar, a higher-order brain region, creates visual perceptions. They delivered controlled electrical pulses to the pulvinar of a patient with epilepsy who had implanted monitoring electrodes. The stimulation also changed brain activity patterns across multiple visual and temporal regions. These results suggest the pulvinar could be a promising target for devices that restore vision.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world use