All authors
Claude Skills by VectorSpaceLab
github.com/VectorSpaceLab6,028 skills13 installs7,823 views
- Frequency Label MappingBuild deterministic one-to-one target-to-source label mappings from source prediction frequencies.Votes: 0GitHub stars: 247
- Iterative Lm Vp OptimizerRun reduced or full ILM-VP alternating label-remapping and prompt-optimization loops.Votes: 0GitHub stars: 247
- Mapping Explanation DiagnosticsDiagnose visual-prompt label-mapping instability, precision, and concept-overlap explanations.Votes: 0GitHub stars: 247
- Visual Prompt WrapperApply and audit universal visual prompts for frozen source-model adaptation experiments.Votes: 0GitHub stars: 247
- Lora Linear UpdateBuild and train a LoRA linear layer with frozen base weights and low-rank trainable factors.Votes: 0GitHub stars: 247
- Lora Merge LatencyMerge LoRA low-rank factors into a dense weight and verify deployment-equivalent inference.Votes: 0GitHub stars: 247
- Lora Parameter BudgetCompute LoRA trainable parameter budgets and reduction ratios for adapted Transformer projections.Votes: 0GitHub stars: 247
- Lora Recovery ProtocolRun a bounded soft-mode LoRA recovery protocol with executable mechanism checks and source-boundary logging.Votes: 0GitHub stars: 247
- Frozen Source Reprogramming TrainingUse this skill when building a recovery experiment that trains only model-reprogramming parameters around a fixed source model. The skill focuses on deterministic small-data optimization traces proving source immutability, parameter updates, loss reduction, and numeric target metrics.Votes: 0GitHub stars: 247
- Masked Input TransformationUse this skill when converting lower-dimensional target vectors into a frozen source model input space by zero-padding and applying a masked, trainable additive reprogramming pattern. It preserves target coordinates and exposes metadata for downstream training checks.Votes: 0GitHub stars: 247
- Mechanism Alignment EvaluationUse this skill after a model-reprogramming recovery run to check whether the observed metric came from the intended mechanism: frozen source model, updated reprogramming parameters, reduced target loss, compatible target metadata, and optional alignment-distance decrease.Votes: 0GitHub stars: 247
- Source To Target Output MappingUse this skill when converting frozen source classifier outputs into target-class probabilities through non-overlapping many-to-one label groups or a small trainable linear head. It is appropriate for model reprogramming recovery when source weights must remain unchanged.Votes: 0GitHub stars: 247
- Vpt Evaluation ReportingCompute VPT recovery accuracy, parameter efficiency, and mechanism report fields.Votes: 0GitHub stars: 247
- Vpt Freeze Scope AuditorAudit VPT trainable scope so only prompts and heads change while backbone stays frozen.Votes: 0GitHub stars: 247
- Vpt Prompt InjectionBuild shallow or deep visual prompt token insertions for VPT-style Transformer inputs.Votes: 0GitHub stars: 247
- Vpt Prompt Tuning LoopRun a bounded tiny VPT optimization loop with frozen backbone and trainable prompts.Votes: 0GitHub stars: 247
- Actor Critic Policy IterationApply policy-iteration style actor updates and verify objective improvement under compatible gradients.Votes: 0GitHub stars: 247
- Compatible Advantage CriticFit and validate compatible advantage critics whose residuals are orthogonal to policy score features.Votes: 0GitHub stars: 247
- Finite Mdp Policy GradientCompute exact policy-gradient theorem checks for finite discounted MDPs with differentiable softmax policies.Votes: 0GitHub stars: 247
- Policy Gradient Recovery HarnessRun a complete mechanism-faithful recovery experiment for the Sutton et al. policy-gradient theorem paper.Votes: 0GitHub stars: 247
- Actor Learner ProtocolValidate IMPALA actor learner unrolls with behavior and learner policy metadata for V-trace recovery.Votes: 0GitHub stars: 247
- Recovery Evaluation HarnessRun a bounded soft-mode IMPALA V-trace recovery harness and emit validator-compatible evidence.Votes: 0GitHub stars: 247
- Vtrace Actor Critic UpdateRun a deterministic V-trace actor-critic optimizer step with value loss and policy-gradient diagnostics.Votes: 0GitHub stars: 247
- Vtrace Target ComputationCompute IMPALA V-trace targets, clipped ratios, and policy-gradient advantages from off-policy trajectories.Votes: 0GitHub stars: 247
- Batched Environment LayoutBuild and validate independent batched robotics environments for Isaac Gym style reduced recovery experiments.Votes: 0GitHub stars: 247
- Parallel Step PipelineExecute a deterministic batched simulation policy reward reset loop for Isaac Gym mechanism recovery.Votes: 0GitHub stars: 247
- Ppo Recovery UpdateRun a deterministic PPO style clipped scalar update over batched rollout evidence for recovery validation.Votes: 0GitHub stars: 247
- Tensor Api Buffer ContractValidate direct state action observation and reward buffer contracts for Isaac Gym style pipelines.Votes: 0GitHub stars: 247
- Ess Policy DistanceCompute P3O effective sample size and derive adaptive clipping and KL coefficients from replay policy probabilities.Votes: 0GitHub stars: 247
- P3o Surrogate LossBuild decomposed P3O surrogate loss terms with clipped replay gradients and behavior-to-target KL regularization.Votes: 0GitHub stars: 247
- Reduced Recovery HarnessExecute a bounded soft-mode proxy experiment that validates P3O mechanisms with generated skills and numeric training evidence.Votes: 0GitHub stars: 247
- Sequential Replay ProtocolModel P3O iteration ordering by appending current rollouts before bounded sequential replay mini-batch updates.Votes: 0GitHub stars: 247
- Mixed Exploration SchedulerAssign mixed Gaussian exploration scales across parallel actors and produce reproducible noisy bounded actions.Votes: 0GitHub stars: 247
- Pql Parallel TopologyBuild and validate the process topology for Parallel Q-Learning with actors, replay, policy learning, and value learning.Votes: 0GitHub stars: 247
- Pql Reduced Recovery HarnessRun a mechanism-faithful reduced Parallel Q-Learning proxy experiment with replay, mixed exploration, and optimizer evidence.Votes: 0GitHub stars: 247
- Speed Ratio Replay DiagnosticsCompute speed-ratio and replay-overwrite diagnostics for massively parallel off-policy Q-learning configurations.Votes: 0GitHub stars: 247
- Ppo Clipped Surrogate ObjectiveCompute and validate PPO clipped probability-ratio surrogate losses and diagnostics for policy update recovery or implementation.Votes: 0GitHub stars: 247
- Ppo Minibatch Update LoopRun a deterministic reduced PPO minibatch update loop with frozen old log probabilities and auditable optimizer traces.Votes: 0GitHub stars: 247
- Ppo Recovery EvaluationValidate reduced PPO recovery results for target consistency, source-boundary compliance, numeric metrics, and mechanism-faithful proxy evidence.Votes: 0GitHub stars: 247
- Ppo Trajectory Advantage EstimationCompute bootstrapped returns and generalized advantage estimates for fixed-horizon PPO rollout batches.Votes: 0GitHub stars: 247
- Proxy Recovery EvaluatorEvaluate soft-mode REINFORCE proxy recoveries for target consistency, numeric metrics, and mechanism-faithful evidence.Votes: 0GitHub stars: 247
- Reinforce Training LoopRun bounded REINFORCE stochastic-policy training loops with sampled actions, scalar rewards, baselines, and optimizer traces.Votes: 0GitHub stars: 247
- Score Function EstimatorCompute REINFORCE score-function update terms for sampled stochastic policy actions with scalar rewards and baselines.Votes: 0GitHub stars: 247
- Maximum Entropy ObjectiveCompute SAC maximum entropy objective diagnostics for rewards, log probabilities, discounts, and temperature-scaled entropy bonuses.Votes: 0GitHub stars: 247
- Sac Recovery HarnessExecute a bounded reduced SAC recovery experiment that combines generated objective, backup, and actor-update skills into validation artifacts.Votes: 0GitHub stars: 247
- Soft Bellman BackupConstruct SAC soft state values, Q targets, and Bellman residuals from replay rewards, log probabilities, and critic estimates.Votes: 0GitHub stars: 247
- Stochastic Actor UpdateRun a deterministic SAC-style stochastic actor update with reparameterized Gaussian action, log probability, and policy-gradient diagnostics.Votes: 0GitHub stars: 247
- Atomic Loss TrainingBuild and optimize the finite-atom APT contrastive loss with proposal posterior correction.Votes: 0GitHub stars: 247
- Proposal Posterior TransformApply APT proposal-prior log-density corrections to candidate posterior scores for likelihood-free inference.Votes: 0GitHub stars: 247
- Recovery EvaluationEvaluate soft-mode APT recovery records for source-boundary, metric, and mechanism-faithfulness checks.Votes: 0GitHub stars: 247