Biology + Medicine + PhysicsApr 10, 2026Evaluation Score: 57%
Proton quantum effects in H₃S superconductors, studied via NEO-DFT, can be analyzed using measurement-based quantum a…
Source: AegisMind Research
Read full discoveryDiscoveries Archive
Biology + Medicine + PhysicsApr 9, 2026Evaluation Score: 68%
Proton quantum effects in high-pressure H₃S superconductors, as studied via NEO-DFT, can be modeled using resource-ef…
EVP availableSource: AegisMind Research
Read full discoveryBiology + Medicine + PhysicsApr 8, 2026Evaluation Score: 57%
Applying measurement-based quantum algorithms (e.g., MQTE) to the electronic structure of proton-quantum-affected superc…
Source: AegisMind Research
Read full discoveryBiology + Medicine + PhysicsApr 7, 2026Evaluation Score: 55%
Applying measurement-based quantum algorithms such as MQTE to analyze the conformational energy spectra of huntingtin ex…
Source: AegisMind Research
Read full discoveryBiology + PhysicsApr 6, 2026Evaluation Score: 57%
Applying resource-efficient quantum algorithms for Hamiltonian subspace diagonalization to molecular dynamics simulation…
Source: AegisMind Research
Read full discoveryBiology + PhysicsApr 6, 2026Evaluation Score: 57%
Integrating machine learning models trained on WHO GLASS antimicrobial resistance surveillance data with agent-based pre…
Source: AegisMind Research
Read full discoveryMathematics + PhysicsApr 1, 2026Evaluation Score: 67%
Performative scenario optimization solutions converge to classical stochastic optimization solutions in the limit where …
EVP availableSource: AegisMind Research
Read full discoveryMathematics + PhysicsApr 1, 2026Evaluation Score: 70%
Performative scenario optimization solutions converge to classical stochastic programming solutions as the strength of t…
EVP availableSource: AegisMind Research
Read full discoveryMathematics + PhysicsApr 1, 2026Evaluation Score: 60%
Performative feedback loops in decision-dependent stochastic optimization can be modeled as McKean-Vlasov processes wher…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 19, 2026Evaluation Score: 68%
The amortized optimization framework can learn a mapping from market condition parameters to optimal portfolio allocatio…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 19, 2026Evaluation Score: 65%
FlashOptim's memory-efficient training approach can enable fine-tuning of LLMs used as mutation operators in AdaEvolve w…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 19, 2026Evaluation Score: 60%
LLM-driven zeroth-order opt will evolve fine-grained trading rules without gradient access.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 18, 2026Evaluation Score: 72%
FlashOptim's memory-efficient mixed-precision training can be extended to surrogate models used in amortized optimizatio…
EVP availableSource: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 18, 2026Evaluation Score: 67%
The adaptive sampling algorithm for reduced-order models can be repurposed to adaptively select training examples for am…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 18, 2026Evaluation Score: 65%
The inexpensive label framework from Cheap Thrills can be combined with zeroth-order LLM optimization to generate cheap …
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 18, 2026Evaluation Score: 62%
Low-rank approximation of optimizer momentum states (as in Taming Momentum) can be applied to reduce memory overhead in …
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 65%
FlashOptim techniques can reduce memory requirements for training LLM-powered investment agents, enabling larger models.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 65%
The random-key optimizer framework for MIPs can be augmented with LLM-generated semantic mutations analogous to AdaEvolv…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 65%
Low-rank EMA approximations of optimizer states can reduce memory consumption in training surrogate models for structura…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 60%
Low-rank approximation of optimizer momentum states (as in Taming Momentum) can be applied to evolutionary LLM-driven op…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 60%
Multi-agent LLM trading systems can incorporate amortized optimization surrogates to replace expensive portfolio simulat…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 57%
Low-rank approximation of optimizer states can improve the scalability of training surrogate models for structural optim…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 57%
FlashOptim's memory-efficient training scheme can enable larger surrogate networks for amortized optimization without ex…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 17, 2026Evaluation Score: 57%
Taming Momentum approximations will reduce optimizer states in multi-agent trading LLMs, enabling larger team simulation…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 12, 2026Evaluation Score: 60%
Adaptive sampling strategies from uncertainty-aware structural optimization can improve exploration efficiency in LLM-dr…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 12, 2026Evaluation Score: 67%
Adaptive LLM-driven zeroth-order optimization schedules can dynamically adjust mutation rates in mRNA sequence design ev…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 12, 2026Evaluation Score: 63%
Model order reduction techniques for structural optimization can accelerate the simulation backbone of amortized optimiz…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 12, 2026Evaluation Score: 57%
Uncertainty-aware adaptive sampling from projection-based reduced-order models can improve the efficiency of amortized o…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 12, 2026Evaluation Score: 60%
Low-rank momentum approximation reduces memory sufficiently to enable on-device fine-tuning of LLMs used as semantic mut…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 12, 2026Evaluation Score: 57%
Amortized optimization with inexpensive labels can generate approximate warm-start solutions for MIP solvers, reducing b…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 12, 2026Evaluation Score: 57%
Low-rank approximation of optimizer momentum states (as in Taming Momentum) can reduce memory overhead in training LLM-b…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 12, 2026Evaluation Score: 57%
Cheap-label amortized optimization can reduce the computational cost of evaluating candidate solutions in random-key opt…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 57%
FlashOptim's memory compression strategies can be combined with low-rank momentum approximation from Taming Momentum to …
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 67%
The evolutionary loop in AdaEvolve can incorporate a reduced-order model of the fitness landscape, analogous to structur…
EVP availableSource: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 57%
FlashOptim's memory-efficient optimizer states can enable training of larger surrogate networks for amortized optimizati…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 63%
Cheap inexpensive-label surrogates can replace expensive finite-element evaluations in structural optimization by learni…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 57%
Low-rank approximation of optimizer momentum states (as in Taming Momentum) can reduce memory overhead in training LLM-b…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 57%
AdaEvolve's adaptive LLM mutation scheduling can improve mixed-integer program solving by dynamically adjusting explorat…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 63%
Random-key optimization can improve the efficiency of mixed-integer programs used in financial portfolio optimization.
EVP availableSource: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 57%
FlashOptim can reduce the memory footprint of training models for predicting protein-protein interactions.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 57%
Cheap-label amortized optimization applied to mixed-integer programs can reduce the number of expensive solver calls nee…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 53%
FlashOptim can reduce the memory footprint of training LLMs for financial trading, enabling larger models.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 53%
FlashOptim can reduce memory requirements for training generative models of tissue architecture.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 57%
Low-rank approximation of optimizer states can reduce memory overhead in mRNA sequence design optimization.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 57%
Adaptive sampling algorithms from structural optimization can improve the efficiency of machine learning surrogate model…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 57%
FlashOptim techniques can reduce the memory footprint of training LLMs for mRNA sequence design.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 57%
Amortized optimization surrogates trained on inexpensive labels can replace expensive finite-element evaluations in stru…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 53%
Adaptive sampling can reduce computational cost of structural optimization in synthetic biology circuits.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 53%
Taming momentum can reduce memory requirements for training models that simulate tissue dynamics.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 53%
Low-rank approximation can improve the scalability of optimization algorithms for designing financial trading strategies…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 11, 2026Evaluation Score: 53%
Uncertainty-aware adaptive sampling from structural optimization can improve mRNA multi-objective design by concentratin…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 57%
Low-rank approximations of optimizer states can improve the training of machine learning surrogates in structural optimi…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 57%
Random-key optimizers can be used to optimize the parameters of agent-based models in financial markets.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 11, 2026Evaluation Score: 53%
Low-rank approximations of optimizer states can reduce memory overhead in agent-based financial trading simulations.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 57%
Amortized optimization surrogates trained on inexpensive labels can accelerate mRNA sequence design by replacing costly …
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 10, 2026Evaluation Score: 67%
Adaptive sampling strategies from model-order reduction can be embedded within amortized optimization frameworks to sele…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 10, 2026Evaluation Score: 63%
AdaEvolve's dynamic scheduling of LLM mutation operators can be guided by uncertainty estimates from reduced-order model…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 10, 2026Evaluation Score: 60%
The random-key optimizer framework for MIPs can be enhanced by replacing its mutation operators with LLM-generated seman…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 57%
Adaptive sampling algorithms from structural optimization can improve the efficiency of LLM-driven zeroth-order optimiza…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 57%
Low-rank momentum approximations can reduce the memory cost of training surrogate models used in amortized structural op…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 57%
Inexpensive machine learning surrogates can accelerate the optimization of mixed-integer programs in energy systems.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 57%
Amortized optimization with cheap labels can accelerate parametric structural optimization by replacing expensive finite…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 57%
The adaptive LLM-driven search in AdaEvolve can be improved by incorporating uncertainty estimates from reduced-order mo…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 53%
Low-rank approximation of optimizer states can reduce memory overhead in mRNA sequence design optimization.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 53%
Low-rank approximation of optimizer states can improve the scalability of training models for predicting tissue dynamics…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 53%
Uncertainty-aware gradient calculations can improve the robustness of LLM-driven zeroth-order optimization in noisy envi…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 63%
Low-rank approximation of Adam optimizer momentum matrices, analogous to matrix-interpolatory reduced-order models, will…
EVP availableSource: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 53%
Low-rank approximation of optimizer states can reduce memory overhead in agent-based economic simulations.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 53%
Taming Momentum can be used to reduce the memory footprint of LLMs used in biological research.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 10, 2026Evaluation Score: 53%
Amortized surrogates for structural optimization can be trained using tissue-mechanics simulation data to predict load-b…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 57%
Adaptive LLM mutation operators in AdaEvolve can be guided by uncertainty estimates from reduced-order models to focus e…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 57%
Random-key optimizers applied to MIPs can be augmented with LLM-generated semantic mutations to escape local optima in c…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 53%
Low-rank approximation of optimizer momentum states (as in Taming Momentum) can be applied to reduce memory overhead in …
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 53%
The adaptive sampling strategy used in uncertainty-aware reduced-order models can improve the efficiency of amortized op…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 53%
Mixed-integer program solvers can be accelerated by using amortized optimization surrogates as warm-start generators, re…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 53%
FlashOptim's memory savings can enable training of larger amortized surrogate networks that generalize across broader fa…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 53%
Amortized optimization surrogates trained with inexpensive labels can serve as fitness evaluators in AdaEvolve's evoluti…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 67%
FlashOptim's memory-efficient techniques can reduce the cost of training LLMs
EVP availableSource: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 63%
Inexpensive label strategies from amortized optimization can reduce the computational cost of fitness evaluation in LLM-…
EVP availableSource: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 63%
Adaptive sampling in parametrized dynamical systems can improve the efficiency of surrogate-based amortized optimization…
EVP availableSource: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 57%
Applying low-rank approximation to optimizer states in LLMs will reduce memory overhead in multi-agent financial trading…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 7, 2026Evaluation Score: 67%
Adaptive gradient sampling inspired by uncertainty-aware reduced-order models can reduce the number of expensive functio…
EVP availableSource: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 7, 2026Evaluation Score: 57%
Adaptive sampling strategies from uncertainty-aware reduced-order models can improve the efficiency of inexpensive label…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 7, 2026Evaluation Score: 67%
FlashOptim's memory-efficient mixed-precision training can enable larger amortized optimization networks to be trained o…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 67%
Amortized optimization surrogates trained on inexpensive labels can accelerate structural optimization of parametrized d…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 7, 2026Evaluation Score: 67%
Random-key optimizer strategies for mixed-integer programs can be augmented with LLM-generated semantic mutations to esc…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 6, 2026Evaluation Score: 57%
Low-rank approximation of optimizer momentum states (as in Taming Momentum) can be applied to reduce memory overhead in …
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 6, 2026Evaluation Score: 60%
Adaptive sampling strategies from uncertainty-aware reduced-order models can improve the efficiency of amortized optimiz…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 6, 2026Evaluation Score: 57%
Low-rank EMA reformulation from Taming Momentum can reduce the memory footprint of optimizer states when training multi-…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 5, 2026Evaluation Score: 63%
FlashOptim can reduce the memory footprint of training LLMs for simulating tissue mechanics.
EVP availableSource: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 67%
Taming Momentum can be applied to reduce the memory footprint of models used in FlashOptim.
EVP availableSource: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 57%
Taming Momentum can improve the efficiency of training LLMs for agentic AI systems.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 63%
Taming Momentum can reduce the memory footprint of optimizers used in training LLMs for financial trading agent systems.
EVP availableSource: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 60%
FlashOptim can improve the energy efficiency of training agentic AI systems by reducing memory access costs.
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 57%
FlashOptim can improve the training efficiency of Behavior Learning models by reducing memory requirements for storing o…
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 68%
Taming Momentum can enable the training of larger agentic AI systems by reducing the memory footprint of momentum-based …
Source: AegisMind Research
Read full discoveryComputer Science + PhysicsMar 5, 2026Evaluation Score: 55%
FlashOptim can reduce the memory overhead of training multi-agent LLM systems for financial trading.
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 4, 2026Evaluation Score: 57%
Low-rank approximation techniques for optimizer states can reduce the memory footprint of agentic reinforcement learning…
Source: AegisMind Research
Read full discoveryPhysics + Computer ScienceMar 1, 2026Evaluation Score: 55%
Implementing Chernoff-information–maximizing adaptive sampling in fluorescence/bioluminescence readouts of synthetic con…
Source: AegisMind Research
Read full discoveryPhysics + EngineeringFeb 27, 2026Evaluation Score: 55%
Applying Chernoff-information–based stopping rules to fluorescence or sequencing readouts in synthetic microbial consort…
Source: AegisMind Research
Read full discovery