Claims Browser

Anthropic has asked prospective employees whether they are comfortable with the possibility that company stock could crash to zero

"Anthropic has sometimes asked prospective employees whether they are comfortable with the possibility that the price of the company stock could potentially crash to zero"

Thomson Reuters has pulled back on its Claude usage

"Thomson Reuters has become the latest company to pull back on its Claude usage"

There is a race between fantasy and reality in AI company valuations and deployment

"The race between fantasy and reality continues"

The regularized Lee-Sidford walk achieves warm-start mixing in Õ((d²+dL²R²)log(w/δ)) steps for polytopes given by n inequalities and a convex L-Lipschitz potential

"For a polytope given by $n$ inequalities and a convex $L$-Lipschitz potential, this yields warm-start mixing in $\widetilde O((d^{2}+dL^{2}R^{2})\log(w/δ))$ steps for the regularized Lee--Sidford walk."

The log-det walk for spectrahedra with n×n blocks achieves mixing in Õ((ψ*nd+dL²R²)log(w/δ)) steps, where ψ* measures matrix leverage

"For a spectrahedron with $n\times n$ blocks, the log-det walk mixes in $\widetilde O((ψ^\star nd+dL^{2}R^{2})\log(w/δ))$ steps, where $ψ^\star$ measures matrix leverage."

Exact-metric Metropolis-adjusted Dikin walks can be analyzed by keeping the proposal determinant and reverse quadratic form together, with their leading uncentered terms canceling in the logarithmic acceptance ratio

"We analyze exact-metric, Metropolis-adjusted Dikin walks by keeping the proposal determinant and reverse quadratic form together. Their leading uncentered terms cancel in the complete logarithmic acceptance ratio, leaving centered fluctuations that can be controlled with second-order tools."

OBABO and BAOAB schemes cannot achieve square-root of condition number acceleration for sampling with any fixed choice of step size and friction based only on curvature bounds and dimension

"We prove that no fixed choice of step size and friction, based only on the curvature bounds and the dimension, achieves this acceleration: total variation mixing time lower bounds for OBABO show that ballistic cold-start mixing fails uniformly over the smooth strongly convex class"

The optimal condition-number dependence of cold-start total variation mixing for fixed-parameter OBABO tunings is linear (not square-root) up to logarithmic factors

"among fixed-parameter OBABO tunings, the optimal condition-number dependence of cold-start total variation mixing over this class is linear, up to logarithmic factors"

A complementary upper bound of O(κ) steps exists for a fixed-parameter OBABO tuning, up to logarithmic factors

"we prove a complementary upper bound of $O(κ)$ steps, up to logarithmic factors, for a fixed-parameter OBABO tuning"

Fixed-parameter acceleration is impossible for the left-endpoint exponential integrator, as demonstrated by direct Gaussian calculation

"A direct Gaussian calculation also rules out fixed-parameter acceleration for the left-endpoint exponential integrator"

Interactions alone may not suffice to bring bio-inspired swarms to a desired aggregate configuration within a prescribed time horizon

"In bio-inspired swarms, however, interactions alone may not suffice to bring the population to a desired aggregate configuration within a prescribed time horizon"

The Schrödinger bridges framework can be applied to finite-horizon minimum-energy collective steering for inertial swarms subject to stochastic disturbances

"Our formalism is rooted in the theory of Schrödinger bridges, which has inspired contemporary developments spanning statistical inference, biological modeling, stochastic control, and generative learning."

Optimal control for swarm steering can dynamically exploit or counteract interaction forces depending on whether they are favorable or adversarial to the steering task

"Numerical examples show that the obtained optimal control (corrective drift) can dynamically exploit or counteract the interaction forces, depending on whether the latter are favorable or adversarial to the steering task."

20-70% of sequenced genes in non-model organisms lack characterized functions

"20-70% of sequenced genes lacking characterized functions"

Homo-RAG with evidence weighting parameter lambda=0.50 achieves NDCG@10 of 0.9879 and MRR of 0.99 for gene function prediction

"evidence weighting parameter of lambda=0.50 improves NDCG@10 to 0.9879 and MRR to 0.99"

Homo-RAG retrieves relevant evidence for 99.33% of queries in gene function prediction tasks

"retrieving relevant evidence for 99.33% of queries"

80% of retrieved documents in Homo-RAG are query-exclusive, indicating evidence quality complements rather than replaces retrieval relevance

"80% of the retrieved documents are query-exclusive, indicating that evidence quality complements rather than replaces retrieval relevance"

Homo-RAG is established as a practical and robust framework for reliable, evidence-grounded gene function prediction in understudied organisms

"These findings establish Homo-RAG as a practical and robust framework for reliable, evidence-grounded gene function prediction in understudied organisms"