Menu
Skills dans ce dépôt
SkillsMP a collecté 4 114 skills depuis hiyenwong/ai_collection. Ouvrez un skill pour examiner sa source et ses détails.
hiyenwong/ai_collectionAffichage de 40 skills collectés sur 4 114.
High-performance NTT accelerator design for post-quantum cryptography (PQC). Novel redundant number representation eliminates conditional corrections for Montgomery modulo multiplication and combined subtract-multiply operations.
Langue du texte source : anglais
Methodology for provably optimal generation of non-Gaussian quantum states (squeezed cat states) via photon catalysis, characterized using stellar rank formalism — enables systematic comparison of fidelity against theoretical maximum for given non-Gaussian…
Langue du texte source : anglais
Pauli noise symmetry analysis from Lindbladian dynamics methodology. Characterizes gate noise channels by exploiting physical symmetry constraints on Pauli fidelities, enabling efficient noise characterization despite SPAM gauge limitations.
Langue du texte source : anglais
Security framework for analyzing distributed attacks in persistent-state AI control. Addresses attack surfaces created by AI coding agents that ship code iteratively across PRs, including gradual attack distribution, monitor evasion strategies, and stateful…
Langue du texte source : anglais
Physical Neural Networks (PNNs) require nonlinearity, signal amplification, and suppression for learning. Shows through simulation that nonlinearity alone is insufficient for meaningful computation in physical computing paradigms. Presents physically…
Langue du texte source : anglais
Post-Quantum Cryptography (PQC) integration pattern for Federated Learning (FL) in Internet of Medical Things (IoMT) environments. Covers edge-native orchestration, PQC key establishment, distributed cryptographic processing, and Kubernetes-based scalable…
Langue du texte source : anglais
PQC-HOT framework for post-quantum cryptography implementation in software systems. Analyzes PQC migration through Human, Organisation, and Technology dimensions. Use when implementing post-quantum cryptography, PQC migration, quantum-safe security…
Langue du texte source : anglais
Training Predictive Coding Networks on ImageNet using Equilibrium Propagation. Biologically plausible training framework for PCNs achieving near-backpropagation performance. Use when: (1) Training predictive coding networks at scale, (2) Implementing…
Langue du texte source : anglais
Program-as-Weights (PAW) paradigm for fuzzy function programming. Compile natural-language specifications into compact, locally-executable neural artifacts using parameter-efficient adapters for frozen interpreters. Achieves 32B-level performance at 1/50…
Langue du texte source : anglais
Programmable dissipation methodology for quantum error correction — treating the QEC cycle as a programmable primitive that turns logical noise into a calibrated resource rather than an adversary. One fault-tolerant round induces a logical completely positive…
Langue du texte source : anglais
Anthropic research (Jun 18, 2026) — Project Fetch phase two showing Claude Opus 4.7 autonomously completing robotic quadruped tasks 20x faster than the fastest human team, revealing the progression pattern from "models helpful to humans" → "humans helpful to…
Langue du texte source : anglais
Predict-Reuse-Repair (PRR) runtime for accelerating dynamic sparse attention in long-context LLM decoding. Speculates attention over predicted KV blocks while selection is in flight, then incrementally repairs missed blocks. Reduces per-token decoding latency…
Langue du texte source : anglais
Proximal Wavefunction Optimization (PWO) methodology for training Neural Quantum States using trust-region optimization. Clips probability-ratio changes in amplitude and phase channels, scales NQS training to billion-parameter models without explicit matrix…
Langue du texte source : anglais
Quantum autoencoder (QAE) for compression-driven anomaly detection in brain MRI data - angle encoding into quantum states, variational encoder-decoder with trash qubits, achieving 0.95 slice-level ROC-AUC.
Langue du texte source : anglais
Practical countermeasure methodology against QKD attacks exploiting detection efficiency mismatch, including time-shift attacks. Uses temporal filtering and detector calibration to ensure uniform detection probability across all degrees of freedom, preventing…
Langue du texte source : anglais
Symmetry exploitation methodology for quantum reservoir computing - observable-orbit completion aligning encoding, dynamics, measurement, and readout for symmetric tasks
Langue du texte source : anglais
Recursive Evidence Replay as LLM Harness for Long-Context Reasoning. Training-free inference method that uses model-internal relevance signals to construct query-conditioned evidence pool and replays before final generation, improving long-context reasoning…
Langue du texte source : anglais
ReContext framework — recursive evidence replay for long-context reasoning, mapping LLM attention to associative memory mechanisms from neuroscience
Langue du texte source : anglais
Distributed containment of compromised CPS agents using repulsive cages — Stackelberg game framework for UAV swarm security via collision-avoidance exploitation. Use when: designing secure multi-agent systems, UAV swarm safety, cyber-physical security,…
Langue du texte source : anglais
Resilient output containment control for heterogeneous multi-agent systems under actuator cyber-attacks. Two-layer adaptive control architecture combining virtual-actuator reconfiguration with network-level adaptive protocols. Handles undisclosed leader…
Langue du texte source : anglais
Framework for distributionally robust regret optimization (DRRO) in stochastic control systems, combining robust control with regret minimization under uncertainty. Use when designing control systems with ambiguous disturbances, robust LQR, or optimizing…
Langue du texte source : anglais
Finite combinatorial analogue of Gödel's incompleteness theorems within Boolean K-SAT. Proves self-referential hardness exhibits physical invariance precluding quantum shortcuts due to necessity of global semantic analysis, and delineates scaling bottleneck…
Langue du texte source : anglais
Complex dynamics in the Sherrington-Kirkpatrick (SK) game methodology — game-theoretic foundation for adaptive learning in disordered many-player systems with random payoff matrices. Generalizes the SK spin-glass model to game theory with random-field bias,…
Langue du texte source : anglais
Microsecond-precision sound localization emerges from slow equilibrium dynamics. ITD represented as stable equilibrium of neural population dynamics rather than classical Jeffress place-coding framework.
Langue du texte source : anglais
Free probability framework for analyzing stationary covariance spectra in discrete-time non-normal random recurrent neural networks. Provides closed-form functional equations for eigenvalue distributions and critical regime behavior.
Langue du texte source : anglais
Free-probability framework for deriving closed functional equations of stationary covariance spectra in discrete-time non-normal random recurrent dynamics, enabling analysis of tail eigenvalues in critical regime.
Langue du texte source : anglais
Short-Term Synaptic Plasticity (STP) stabilizes goal-conditioned dynamics in PFC-inspired reservoir model for multistep goal-directed action planning. Preserves action-relevant goal information under noise with 89.2% success rate vs 49.5% without STP.…
Langue du texte source : anglais
Structure-aware variance reduction methodology for unbiased randomized Hamiltonian simulation. Combines classical variance reduction with randomized product-formula estimators to achieve 70-96% sampling cost reductions in tensor-network simulations. Use when…
Langue du texte source : anglais
Mean-field theory bridging microscale synaptic motifs to macroscale heterogeneous population dynamics in neural networks
Langue du texte source : anglais
Mean-field theory linking microscale synaptic motifs to macroscopic heterogeneous population dynamics in neural networks. Use when studying synaptic-resolution connectomics, second-order motifs, random RNNs with cell types, or heterogeneous population…
Langue du texte source : anglais
Mean-field theory for heterogeneous synaptic motifs in multi-population neural networks. Bridges microscale synaptic connectivity (second-order motifs) to macroscale population dynamics. Activation: synaptic motifs, mean-field theory, heterogeneous dynamics,…
Langue du texte source : anglais
Thermocoherent framework for modeling information flow in neural matter. Heat flow couples to delocalized information flow carried by shared coherence. Use when: thermocoherent effects in neural systems, quantum cognition physical basis, relational resources…
Langue du texte source : anglais
Compute controlled invariant sets for linear discrete-time systems using trajectory-based approach with convex feasible points. Use for MPC design, control invariant set computation, feasibility guarantees, and discrete-time system analysis. Keywords:…
Langue du texte source : anglais
Role-typed credit assignment framework for agentic reinforcement learning that addresses outcome-only credit blind spots
Langue du texte source : anglais
Brain foundation models (BFMs) lose third-order statistics (co-skewness) during pretraining - variance allocation problem where billion-parameter models predict cognition worse than linear FC. Solution: preserve co-skewness subspace. Activation: brain…
Langue du texte source : anglais
Self-specializing Vision-Language Model agent for physics-grounded transmon chip calibration. Uses zero-weight-update online adaptation via human-readable device notes, gradient-free strategy refinement with paired-snapshot accept gate, and physics-grounded…
Langue du texte source : anglais
Magic Rényi entropy methodology unifying quantification of nonstabilizerness and non-Gaussianity across spins, bosons, and fermions using conformal field theory analysis.
Langue du texte source : anglais
PREDIKTOR: Patient-centered multi-view framework aligning personalized knowledge graphs with gene-level perturbation representations for clinical drug response prediction. Combines DysRegNet GRN construction, DrugBank integration, GNN encoding, LINCS L1000…
Langue du texte source : anglais
Comprehensive review of the AI-Quantum Information interface — covering AI for quantum systems (measurement, algorithm discovery, hardware stabilization) and quantum for AI (algorithmic speedups, expressivity, trainability, generalization).
Langue du texte source : anglais
Automated framework for synthesizing inter-code logical Clifford (CNOT/CZ) gates between arbitrary CSS quantum error-correcting codes using chain maps — recovers known transversal constructions and finds new low-depth solutions for heterogeneous QEC…
Langue du texte source : anglais