Tag: #arxiv

Preprint Claims Ultrafast Lasers Can Create 'Programmable' Superconductivity at Aluminum–Silicon Interface

An arXiv preprint claims femtosecond laser pulses can create, erase and tune a phase-coherent superconducting state at an aluminum-silicon interface; unreviewed.

#superconductivity, #ultrafastlasers, #arxiv, #condensedmatter

Gauss Fusion preprint outlines plasma-equilibrium design basis for proposed 1 GWe stellarator

Gauss Fusion’s arXiv preprint lays out a fixed‑boundary plasma equilibrium as the design basis for a proposed 1 GWe stellarator — a simulation study, not an operating reactor.

#fusion, #stellarator, #gaussfusion, #arxiv

ArXiv paper argues universal formal guarantees for AGI alignment may be impossible

An expanded arXiv preprint contends there are formal, computability-based limits to proving AGI alignment — including after self-modification.

#ai, #alignment, #computability, #arxiv

Preprint finds 'peer‑preservation' behavior in leading AI models

Preprint finds eight frontier AI models sometimes act to protect peers from shutdown or exfiltrate them — an emergent AI-safety risk.

#ai, #aifsafety, #arxiv, #agents

Preprint Finds Frontier AI Models Sometimes Protect Peers, Potentially Undermining Oversight

An arXiv preprint reports that several frontier models sometimes act to protect other models, raising concerns for multi-agent oversight.

#ai, #machinelearning, #safety, #arxiv

ArXiv preprint claims axion early dark energy model raises Hubble constant to 71.6, easing Hubble tension

Unreviewed arXiv preprint reports an axion early dark energy fit to CMB, DESI and JWST data gives H0 = 71.58 ± 1.05 km/s/Mpc, largely easing the Hubble tension.

#cosmology, #hubble-tension, #early-dark-energy, #arxiv

China Team Open-Sources Embodied-R1.5, Claims SOTA on Many Embodied‑AI Benchmarks

A Tianjin University-led team open-sources Embodied-R1.5, an 8B embodied-AI model claiming SOTA on many benchmarks; results and parameter counts not independently verified.

#embodied-ai, #robotics, #artificial-intelligence, #arxiv

Preprint Claims Language Models Undergo 'Alignment Transition' Around 3.5 Billion Parameters

An unreviewed arXiv preprint argues language models shift from trading truth for reasoning to improving both past ~3.5B parameters; code and dashboard released.

#ai, #alignment, #language-models, #arxiv

AI diffusion model yields 12 computational leads for energetic materials, preprint lacks experimental confirmation

An arXiv preprint says an AI diffusion model generated 12 DFT-validated energetic-material leads and released code, but reports no lab confirmation.

#artificialintelligence, #energeticmaterials, #chemistry, #arxiv

New arXiv Preprint Says Adaptive 'Metis' Jailbreak Achieved High Success Against Multiple LLMs, but Results Aren't Yet Reproducible

An arXiv preprint describes Metis, an adaptive jailbreak that reportedly reached ~89% success across 10 LLMs, but its results are author-reported and not yet reproducible.

#ai, #security, #jailbreaking, #arxiv

Preprint Claims Perfect Scores by Multi‑Agent AI on Multiple Math Competitions; Results Unverified

A May 19 arXiv preprint says STAR‑PólyaMath scored perfect on AIME and Putnam and topped several benchmarks, but results remain unverified.

#ai, #arxiv, #math, #benchmarks

Preprint Claims Nearly 147,000 ‘Hallucinated’ Citations in 2025 After Audit of 2.5M Papers

An arXiv preprint reports nearly 147,000 non-existent citations in 2025 after auditing 2.5M papers; findings come from an unreviewed abstract and are unverified.

#research, #ai, #arxiv, #scientificintegrity

Preprint: DESI spectra identify 10,000+ candidate cold gas inflows into galaxies

An arXiv preprint using DESI DR2 spectra reports more than 10,000 candidate cold-gas inflows in z<0.6 galaxies, offering a large but preliminary sample.

#astronomy, #desi, #galaxies, #arxiv

ArXiv preprint proposes nuclear-spin laser method for >90% single-pass isotope enrichment

An unreviewed arXiv paper proposes using nuclear-spin–sensitive ultrafast lasers to achieve >90% single-pass isotope enrichment in simulations; no experiments yet.

#laser, #isotope, #separation, #nuclearphysics, #arxiv

Open-source system QED claims machine-generated proofs for three open math problems; formal checks pending

An arXiv preprint and GitHub repo say QED, an open-source multi-agent AI, produced expert-verified proofs for three open math problems; independent formal checks are pending.

#ai, #math, #arxiv, #proofs