All authors
Claude Skills by VectorSpaceLab
github.com/VectorSpaceLab6,028 skills13 installs7,823 views
- Rlaif Evaluation MetricsCompute RLAIF AI-labeler alignment, pairwise win rate, harmless rate, and proxy target consistency metrics.Votes: 0GitHub stars: 247
- Rlaif Preference LabelingBuild RLAIF pairwise AI-feedback labels from label-token logits, optional chain-of-thought rationales, and swapped-order position-bias mitigation.Votes: 0GitHub stars: 247
- Rlaif Reward ModelingTrain and validate tiny soft-label pairwise reward models that distill RLAIF AI preference distributions with cross-entropy.Votes: 0GitHub stars: 247
- Toxigen Alice DecodingApply ToxiGen ALICE classifier-in-the-loop weighted beam scoring for adversarial or detoxifying generation experiments.Votes: 0GitHub stars: 247
- Toxigen Classifier RecoveryRun a bounded ToxiGen-style classifier update and report validator-compatible before-after loss, AUC, and parameter evidence.Votes: 0GitHub stars: 247
- Toxigen Human Eval MetricsCompute ToxiGen-style human harm labels, classifier attack flags, and aggregate toxicity validation metrics.Votes: 0GitHub stars: 247
- Toxigen Prompt ProtocolBuild and validate ToxiGen-style balanced identity-label demonstration prompts for implicit toxicity generation and recovery experiments.Votes: 0GitHub stars: 247
- Anchor Patch ProtocolRoute latent samples through anchor image realism or patch realism for few-shot adaptation.Votes: 0GitHub stars: 247
- Cdc Similarity LossCompute cross-domain distance consistency softmax-KL losses for few-shot generator adaptation experiments.Votes: 0GitHub stars: 247
- Diversity Correspondence MetricsCompute diversity and correspondence diagnostics for few-shot image generation recovery experiments.Votes: 0GitHub stars: 247
- Few Shot Adaptation ObjectiveExecute a bounded CDC-regularized few-shot adaptation optimizer step with route-aware realism terms.Votes: 0GitHub stars: 247
- Dual Contrastive LossCompute and inspect a DCL-style contrastive proxy loss with same-latent positives and real-target anti-collapse negatives.Votes: 0GitHub stars: 247
- Intra Lpips Proxy EvaluationCompute a deterministic intra-cluster diversity proxy mirroring intra-LPIPS evaluation for reduced few-shot generation recovery.Votes: 0GitHub stars: 247
- Latent Pair ProtocolValidate DCL latent-paired source/target feature batches and separate real-target negatives for few-shot GAN adaptation recovery.Votes: 0GitHub stars: 247
- Reduced Training HarnessExecute a bounded finite-difference optimizer step for DCL proxy recovery and emit validator-compatible training traces.Votes: 0GitHub stars: 247
- Epsilon Prediction LossCompute and test the DDPM epsilon-prediction mean squared error training objective.Votes: 0GitHub stars: 247
- Forward Diffusion ScheduleBuild DDPM beta schedules and closed-form forward noising coefficients for recovery experiments.Votes: 0GitHub stars: 247
- Reduced Ddpm Recovery HarnessRun a bounded soft-mode proxy DDPM recovery experiment using generated module skills.Votes: 0GitHub stars: 247
- Reverse Sampling PosteriorCompute DDPM predicted x0 and reverse posterior statistics from predicted epsilon.Votes: 0GitHub stars: 247
- Ddpm Adaptation ProtocolBuild shared noised DDPM adaptation batches and reconstruct clean-image predictions for source and adapted denoisers.Votes: 0GitHub stars: 247
- Ddpm Pa Recovery HarnessRun a bounded DDPM-PA proxy optimizer step and emit mechanism-faithful recovery traces and metrics.Votes: 0GitHub stars: 247
- Haar High Frequency EnhancementExtract Haar high-frequency components and compute DDPM-PA high-frequency preservation and detail losses.Votes: 0GitHub stars: 247
- Pairwise Similarity PreservationCompute DDPM-PA pairwise cosine-softmax distributions and adapted-to-source KL preservation loss.Votes: 0GitHub stars: 247
- Fid Activation StatisticsCompute FID-style activation mean/covariance summaries for generated or real feature matrices.Votes: 0GitHub stars: 247
- Fid Ttur Recovery ProtocolAssemble a bounded soft-mode recovery protocol that exercises FID statistics, FID distance, and TTUR update skills.Votes: 0GitHub stars: 247
- Frechet Gaussian DistanceCompute the Fréchet Gaussian distance used by FID with stable numerical handling.Votes: 0GitHub stars: 247
- Ttur Update ScheduleRun deterministic two-time-scale update traces for GAN-style min-max proxy dynamics.Votes: 0GitHub stars: 247
- Adaptive Unet SpecValidate paper-critical guided diffusion UNet architecture flags before claiming architecture-faithful recovery or sampling.Votes: 0GitHub stars: 247
- Classifier Guidance SamplerApply noisy classifier log-probability gradients to reverse diffusion steps with explicit guidance-scale checks.Votes: 0GitHub stars: 247
- Diffusion Noise ScheduleBuild and validate Gaussian diffusion beta schedules and posterior coefficients for guided diffusion recovery experiments.Votes: 0GitHub stars: 247
- Upsampling Evaluation ProtocolPreserve guided diffusion target metadata, proxy declarations, classifier scale, and evaluation protocol consistency.Votes: 0GitHub stars: 247
- Convolutional Spatial SamplingCompute and validate LDM spatial latent-grid sampling plans for high-resolution and dense conditional tasks.Votes: 0GitHub stars: 247
- Cross Attention ConditioningValidate and run deterministic LDM-style cross-attention conditioning between latent features and modality tokens.Votes: 0GitHub stars: 247
- Latent Diffusion ObjectiveExecute and validate a reduced LDM latent noising, epsilon prediction loss, and optimizer-step training primitive.Votes: 0GitHub stars: 247
- Perceptual Latent CompressionValidate LDM autoencoder latent compression contracts, spatial downsampling, regularization metadata, and reconstruction diagnostics.Votes: 0GitHub stars: 247
- Reduced Recovery HarnessBuild executable soft-mode LDM proxy recovery artifacts from generated skills with mechanism checks and validator evidence.Votes: 0GitHub stars: 247
- Bapps 2afc ProtocolEvaluate BAPPS-style two-alternative forced choice perceptual similarity triplets with auditable per-item predictions.Votes: 0GitHub stars: 247
- Linear Calibration RankingFit non-negative layer calibration weights for LPIPS-style distances using bounded 2AFC ranking feedback.Votes: 0GitHub stars: 247
- Lpips Feature DistanceCompute LPIPS-style perceptual patch distances from normalized multi-layer feature differences with optional non-negative calibration weights.Votes: 0GitHub stars: 247
- Lpips Recovery HarnessRun a bounded source-safe LPIPS-style recovery experiment and emit validation-compatible recovery artifacts.Votes: 0GitHub stars: 247
- Cic Contrastive LossCompute CIC normalized skill-transition contrastive loss, logits, margins, and tiny dependency-free training updates.Votes: 0GitHub stars: 247
- Particle Entropy RewardEstimate CIC intrinsic reward using k-nearest-neighbor particle entropy over learned transition embeddings.Votes: 0GitHub stars: 247
- Reduced Cic Recovery HarnessRun a bounded soft-mode CIC proxy experiment that invokes generated batching, contrastive loss, and entropy reward skills.Votes: 0GitHub stars: 247
- Transition Skill BatchingBuild and validate aligned CIC state-transition and continuous skill batches for contrastive intrinsic control recovery experiments.Votes: 0GitHub stars: 247
- Benchmark Property TaxonomyClassify offline RL datasets by D4RL challenge properties and explain benchmark implications from metadata.Votes: 0GitHub stars: 247
- D4rl Score NormalizationCompute and audit D4RL normalized returns from raw, random-baseline, and expert-reference scores.Votes: 0GitHub stars: 247
- Offline Dataset ProtocolValidate fixed offline reinforcement learning transition datasets and summarize D4RL-style provenance before recovery experiments.Votes: 0GitHub stars: 247
- Offline Recovery HarnessExecute a bounded D4RL-style offline recovery proxy with fixed-dataset training and normalized-score evidence.Votes: 0GitHub stars: 247
- Diayn Discriminator Intrinsic RewardCompute DIAYN discriminator log-probabilities, cross-entropy diagnostics, and intrinsic rewards from skill-conditioned state predictions.Votes: 0GitHub stars: 247
- Diayn Maxent Policy UpdateExecute a bounded maximum-entropy DIAYN policy surrogate update with before-after loss and parameter-change evidence.Votes: 0GitHub stars: 247