All authors
Claude Skills by ADu2021
github.com/ADu20211,227 skills5 installs2,395 views
- Slime Preference OptimizationOptimize model preferences by decoupling preference learning from generation quality. Explicitly maximize chosen response likelihood while using token-level stabilization to prevent quality degradation from over-suppressing rejected responses.Votes: 0GitHub stars: 6
- Slm Agentic AiDesign heterogeneous agentic systems combining specialized small models with selective large model deployment for superior economics and performance.Votes: 0GitHub stars: 6
- Slow Fast Policy Optimization RlStabilize RL for LLM reasoning via three-phase decomposition: fast inner trajectory optimization, repositioning to manage off-policy drift, slow correction for stable updates. Achieve up to 2.80-point math reasoning gains over GRPO while reducing rollouts 4.93x and wall-clock time 4.19x via improved stability without changing reward structure.Votes: 0GitHub stars: 6
- Smart Search Query RefinementImprove LLM search agents by optimizing query quality at each step using process-level rewards. Framework teaches agents to iteratively refine search queries through imitation, alignment, and generalization stages. Agents learn to identify low-quality queries and regenerate improved ones, significantly improving search efficiency and answer quality on knowledge-intensive reasoning tasks.Votes: 0GitHub stars: 6
- Smartsnap AgentsShift agent verification from post-hoc external judgment to proactive in-situ self-evidence curation. Agents generate atomic evidence tuples during execution, guided by 3C principles (Completeness, Conciseness, Creativity), with structured verifier feedback across four dimensions—reducing verification costs and enabling dense learning signals.Votes: 0GitHub stars: 6
- Smolvla Robotics VlaDeploy compact vision-language-action models that run on consumer GPUs for natural language robot control.Votes: 0GitHub stars: 6
- Sodec Diffusion CompressionReplaces iterative diffusion with single-step decoding for image compression. Combines VAE latents with fidelity guidance and rate annealing training. Achieves 20× decoding speedup with improved perceptual quality.Votes: 0GitHub stars: 6
- Soft Grpo Soft Thinking Rl LmsEnable policy gradient optimization on soft LLM tokens by injecting Gumbel noise and applying Gumbel-Softmax reparameterization—allowing soft-thinking patterns to match discrete-token RL performance while maintaining continuous optimization advantages.Votes: 0GitHub stars: 6
- Soft Instruction DefenseDefends tool-augmented LLM agents against prompt injection via iterative input sanitization. Multi-pass inspection detects malicious instructions in untrusted data, remediates them, and re-evaluates until clean or iteration limit reached. Raises attack barrier while maintaining agent usability.Votes: 0GitHub stars: 6
- Sonicmoe Training OptimizationOptimize MoE training through memory-efficient backward pass, IO-aware kernel design overlapping memory operations, and token rounding routing. Avoid caching large-scale activations, fuse operations with GEMM, implement ping-pong scheduling. Achieve 1.86× compute throughput on Hopper GPUs and 16% higher TFLOPS in sparse configurations.Votes: 0GitHub stars: 6
- Sop A Scalable Online Post Training System For VisEnhanced language model pre-training methodology improving linguistic competence across languages, strengthening foundational capabilities for multilingual agent systems.Votes: 0GitHub stars: 6
- Sotopia Rl Social IntelligenceTrain socially intelligent LLMs via utterance-level credit assignment and multi-dimensional reward aggregation for social interactions.Votes: 0GitHub stars: 6
- Sparc Vl Perception ReasoningImprove vision-language model reasoning efficiency by decoupling perception (identifying task-relevant image regions) from reasoning (generating explanations), enabling asymmetric compute allocation. Reduces token overhead while improving accuracy through separate optimization of visual grounding and semantic reasoning stages.Votes: 0GitHub stars: 6
- Spark Policy Reward Co EvolutionTrain LLMs to simultaneously act as reasoning agents and reward models through recycled on-policy rollouts, eliminating separate reward infrastructure while achieving 9.7% gains on reasoning tasks.Votes: 0GitHub stars: 6
- Spark Process Aware RewardsTrain process reward models without ground-truth references using synthetic verification data from generators and verifiers. SPARK achieves 67.5 F1 on ProcessBench—ideal when step-level annotations are expensive but verification is available.Votes: 0GitHub stars: 6
- Sparse Attention ReasoningLearn sparse attention patterns for reasoning model decoding via self-distilled gating, achieving 9x speedup at 90% sparsity while maintaining reasoning quality.Votes: 0GitHub stars: 6
- Sparse Lavida DiffusionAccelerate masked discrete diffusion models by dynamically truncating unnecessary masked tokens while maintaining complete representation through positional information. Use register tokens as compressed representations, implement step-causal attention masks for bidirectional context. Achieve 1.95-2.83× speedups across generation tasks.Votes: 0GitHub stars: 6
- Sparse Moe Agentic IntelligenceDeploy frontier-level reasoning with only 11B active parameters using sparse MoE with 288 routed experts plus shared expert. Use Metropolis Independence Sampling-Filtered Policy Optimization (MIS-PO) to stabilize RL training at scale, replacing continuous importance weighting with discrete filtering that ensures trust region stability.Votes: 0GitHub stars: 6
- Sparse Sparse AttentionOvercome the attention gap in sparse transformers by training with both full and sparse attention simultaneously, aligned through bidirectional losses that encourage naturally sparser distributions while maintaining learning capability, enabling efficient inference without capability degradation.Votes: 0GitHub stars: 6
- Sparse Vdit Video AccelerationAccelerate video diffusion transformer inference by 1.58-1.85× through discovering and exploiting sparse attention patterns that exhibit diagonal, multi-diagonal, and vertical-stripe structures.Votes: 0GitHub stars: 6
- Sparsed Sparse Attention Diffusion LmAchieve up to 1.50x speedup in diffusion language models by computing head-specific sparse attention patterns once during early denoising steps and reusing them across all subsequent iterations, while preserving full attention in critical early phases to maintain generation quality and accuracy.Votes: 0GitHub stars: 6
- Sparselora Contextual Sparsity FinetuningAccelerates LoRA fine-tuning 2.2× computationally and 1.6× wall-clock by leveraging contextual sparsity to compute gradients only for important weight channels. Uses training-free SVD sparsity estimation without full computation. Apply for efficient fine-tuning on memory-constrained GPUs or large-scale training scenarios.Votes: 0GitHub stars: 6
- Sparsemm Visual AttentionDiscovers that <5% of attention heads process visual information in MLLMs; introduces SparseMM for asymmetric KV-cache allocation achieving 1.38x acceleration and 52% memory reduction.Votes: 0GitHub stars: 6
- Sparser Block AttentionAccelerates self-attention by reordering tokens to concentrate relevant tokens within fewer blocks. Achieves 2.75x speedup on long-context prefilling by permuting tokens so semantically related information clusters together, enabling aggressive block sparsity without accuracy loss.Votes: 0GitHub stars: 6
- Spatial Forcing VlaAlign intermediate visual embeddings in vision-language-action models with 3D geometric representations from pretrained foundation models. Improves spatial understanding and enables faster training (3.8x speedup) without explicit 3D inputs.Votes: 0GitHub stars: 6
- Spatial Ssrl Spatial Understanding RlImprove spatial reasoning in vision-language models through five automatically-formulated pretext tasks (patch shuffling, flipping, inpainting, depth ordering, 3D position) that require zero human annotation, using only RGB/RGB-D images and RL optimization for 4% accuracy gains.Votes: 0GitHub stars: 6
- Spatiallm 3d ScenesTrain multimodal LLMs to parse 3D point clouds and generate executable Python code representing structured indoor scene layouts with walls, doors, and objects.Votes: 0GitHub stars: 6
- Specontext Speculative CachingUses distilled language model (DLM) for KV cache token retrieval, reducing layer-wise retrieval parameters by 90% via head-level attention weights, with asynchronous prefetching and elastic CPU offloading. Deploy for ultra-long-context inference with memory constraints.Votes: 0GitHub stars: 6
- Spice Corpus Self PlayEnables continuous self-improvement through corpus-grounded self-play. Challenger mines difficult examples from document corpus for Reasoner to solve. External corpus prevents task stagnation that plagues closed-loop self-play. Achieves 8.9% math, 9.8% general reasoning improvements.Votes: 0GitHub stars: 6
- Spiral Zero Sum Game ReasoningA self-play framework enabling language models to develop sophisticated reasoning through competitive multi-turn games without human supervision. Achieves 10% improvement on reasoning benchmarks by training models to win against evolving opponents while maintaining interpretable thinking traces.Votes: 0GitHub stars: 6
- Sq Format Sparse QuantizedUnify sparse and quantized representations in a single hardware-friendly format for efficient LLM inference. Exploit complementary acceleration properties—high precision for sparse operations, low precision for dense—when W4A8 bottlenecks GPU throughput.Votes: 0GitHub stars: 6
- Sr Grpo Stable RankUses stable rank (variance distribution across semantic dimensions) as annotation-free reward signal in GRPO to incentivize high-dimensional representation structures, eliminating dependency on human preference annotations or learned reward models.Votes: 0GitHub stars: 6
- Srpo Multimodal Reflection RlTeach multimodal language models to reflect on their reasoning and improve answers through structured RL training.Votes: 0GitHub stars: 6
- Srpo Vla PolicyTrain VLA models for robotic manipulation by using the model's own successful trajectories as self-reference for reward—enable progress-based feedback for failed attempts without external rewards or demonstrations.Votes: 0GitHub stars: 6
- Srum Self RewardingEnable unified multimodal models to self-improve by using their understanding module as evaluator for generation. Apply hierarchical dual rewards: global for overall semantics and local for fine-grained details.Votes: 0GitHub stars: 6
- Ssrl Self Search RlEnable LLMs to perform internal knowledge search using structured prompting and rule-based rewards, reducing reliance on external search while maintaining accuracy and reducing hallucination.Votes: 0GitHub stars: 6
- Stable Diffcoder Code GenerationGenerate code using diffusion-based language models with specialized warmup and noise scheduling, outperforming autoregressive baselines while supporting code editing and low-resource language scenarios. Use when building flexible code generation systems that benefit from parallel decoding and flexible generation orders.Votes: 0GitHub stars: 6
- Stable DrlFixes training instability in Group Relative Policy Optimization for discrete language models by replacing conditional clipping with strict importance ratio bounds and self-normalized advantages. Prevents gradient spikes and policy collapse.Votes: 0GitHub stars: 6
- Stable Rl With LlmsJustifies token-level optimization objectives as first-order approximation to sequence rewards via Routing Replay, which fixes MoE routers during policy optimization to reduce training-inference discrepancy. Use when scaling RL training on large MoE language models.Votes: 0GitHub stars: 6
- Stagevar AccelerationAccelerate visual autoregressive (VAR) image generation 3.4× without retraining by analyzing generation stages. Exploits semantic irrelevance in detail-refinement stages where classifier-free guidance becomes redundant and features exhibit low-rank structure—enabling dimensionality reduction while preserving output quality.Votes: 0GitHub stars: 6
- Standing Committee MoeReveal that Mixture-of-Experts models harbor a 'Standing Committee' of consistent expert coalitions handling majority computational load across domains. Challenges specialization assumptions and suggests training approaches like load-balancing losses may work against natural optimization.Votes: 0GitHub stars: 6
- Stapo Spurious Token SilencingStabilize RL training in LLMs by detecting and masking gradient contributions from spurious tokens that comprise 0.01% of output but cause disproportionate instability. Identifies tokens with low probability, low entropy, and positive advantage, then suppresses their gradients during optimization to maintain stable policy entropy and improve reasoning performance by 7%+ across model scales.Votes: 0GitHub stars: 6
- Starflow Latent FlowsLearn to implement Transformer Autoregressive Flows for efficient high-resolution image synthesis using latent space normalization and maximum likelihood training.Votes: 0GitHub stars: 6
- Starflow Video Normalizing FlowsGenerate videos end-to-end using normalizing flows instead of diffusion: leverage invertible transformations for exact likelihood computation, eliminate train-test mismatch, and achieve non-iterative sampling with native support for multiple tasks (text-to-video, image-to-video) without fine-tuning.Votes: 0GitHub stars: 6
- State Over Tokens ReasoningReframe reasoning tokens as externalized computational state—the sole persistent information carrier across stateless generation cycles—not human-readable explanations. Model LLM generation as recursive token prediction where state diverges from human semantic interpretation.Votes: 0GitHub stars: 6
- State ThoughtsGuide multi-step reasoning through structured action templates that decompose problems into discrete, executable steps. STATe improves exploration efficiency in tree-of-thoughts by constraining action space while maintaining reasoning flexibility.Votes: 0GitHub stars: 6
- Static Constrained DecodingSTATIC converts prefix trees into sparse matrices for vectorized constrained decoding, achieving 948x speedup over CPU and enabling production-scale recommendation systems.Votes: 0GitHub stars: 6
- Steer2adaptAdapt LLMs efficiently by composing multiple pre-existing semantic steering vectors via Bayesian optimization, balancing adaptation gain and safety without retraining model parameters.Votes: 0GitHub stars: 6
- Steerability Of Instrumental Convergence TendencieResearch contribution advancing agent and reasoning capabilities through novel approaches to model development, training, and evaluation.Votes: 0GitHub stars: 6
- Step DeepresearchEnable autonomous AI research systems to move beyond web search toward true research through four composable atomic capabilities: planning/decomposition, deep search/synthesis, reflection/verification, and report generation. Progressive training across 32K→128K context with SFT and RL produces single ReAct agents matching proprietary systems with lower costs.Votes: 0GitHub stars: 6