In briefGenerating vector-vortex gamma photons via nonlinear Compton scattering
SummaryThis paper proposes a new method to generate vector-vortex gamma photons, whose polarization and orbital angular momentum are nonseparably coupled, using nonlinear Compton scattering in elliptically polarized laser pulses. The authors report that tuning laser ellipticity directs multiphoton absorption to coherently populate orbital angular momentum modes with winding numbers ±ℓ, producing spin-OAM nonseparable photon states. For moderate-intensity linearly polarized lasers, they state the mode-pair concurrence reaches unity for MeV gamma photons.
Yong-Zheng Ren, Mamutjan Ababekri, Jun-Lin Zhou et al.
In briefAnalyzing Solana's Blocks and Transactions: A Statistical Study of a Million Recent Blocks
SummaryThis paper studies how smart contracts (dApps) run on the Solana blockchain, analyzing about one million recent blocks. It reports statistics on transactions per block, execution time units and fees for individual transactions, and the Solana programs invoked. Using the declared read and write sets that Solana requires for each transaction, the authors analyze conflicts and conflict graphs within blocks to assess the network's potential for parallelism. The data and code are open source.
In briefSmartMemory: Using a Memory-Based Agent to Detect On-Chain/Off-Chain Inconsistencies in Smart Contracts
SummaryThis paper proposes SmartMemory, a framework that uses a memory-based agent to detect on-chain/off-chain communication inconsistency (OFCI), which the authors identify as a root cause of vulnerabilities in such applications. It maps diverse contract implementations into a unified business-semantic representation and transfers vulnerability knowledge to find unseen cases, then uses taint analysis to verify candidates. On a dataset of 48 DApps with 81 OFCIs, it reports 80.68% precision and 87.65% recall, and it found 36 previously unknown OFCIs across 325 real-world applications, all confirmed and fixed.
WhyIntroduces a unified OFC inconsistency view and first memory-based agent detection; strong precision/recall on 81 OFCIs and 36 confirmed unknown bugs.
In briefCan AI Earn Profits in Crypto? Measuring the Gap Between Backtests and Real Markets
SummaryThis paper presents a unified benchmark for evaluating AI trading methods, including machine learning, reinforcement learning, LLM-based, and agent-based approaches, in cryptocurrency markets. The evaluation proceeds in three increasingly realistic stages: historical backtesting, exchange-based paper trading, and real-money live trading. This design lets the authors quantify the gap between backtested and realized performance and compare how that gap differs across method classes. The authors also release an open-source system supporting all three stages and a public platform that continuously updates benchmark results.[ ]
In briefReconstructing nuclear scattering phase shifts on a quantum processor with compressed RODEO circuits
SummaryThe paper reconstructs elastic s-wave neutron-proton phase shifts from trapped energy spectra measured on an IBM quantum processor using a compressed RODEO circuit. It introduces a six-cycle factorized geometric-time RODEO method (FG-R6) and compares it with ten-cycle direct RODEO at equal qubit width and shot budgets. FG-R6 reduces median compiled depth and two-qubit-gate count and lowers finite-shot energy uncertainty. The authors argue that circuit performance should be judged through the scattering reconstruction and its energy range, not spectral RMS errors alone.
Myeong-Hwan Mun, Jubin Park, Myung-Ki Cheoun et al.
In briefPlacement-driven reversal of preferred spin alignment in monolayer hBN vacancy pairs
SummaryThis paper studies how two negatively charged boron vacancies in monolayer hexagonal boron nitride align their spins when they interact. Using density-functional theory across seven configurations, the authors find that moving one vacancy by a single lattice vector reverses the preferred collinear spin alignment. The closest pair, 4.3 angstrom apart, favours parallel alignment, while a 7.5 angstrom pair favours antiparallel alignment.
