Menú
Skills en este repositorio
SkillsMP ha recopilado 1228 skills de ADu2021/skillXiv. Abre una skill para revisar su origen y sus detalles.
ADu2021/skillXivMostrando 40 de 1228 skills recopiladas.
Automatically generate high-quality domain-specific fine-tuning datasets from raw documents using adaptive processing and persona-driven synthesis, preserving general capabilities while specializing models.
Idioma del texto original: inglés
Benchmark LLMs on real-world forecasting questions from Metaculus, comparing against human crowds and expert forecasters. Identifies which domains LLMs handle well and where they fall short relative to human intelligence.
Idioma del texto original: inglés
Interpret LLM behavior as locally linear mappings between hidden representations, enabling mechanistic understanding of computation without examining individual weights or attention patterns.
Idioma del texto original: inglés
Understand and mitigate unintended memorization of training data in LLMs by systematizing mechanisms, detection methods, and mitigation strategies across the model lifecycle.
Idioma del texto original: inglés
Enable language models to explore code sandboxes to solve diverse tasks like mathematics and long-context reasoning without additional training, with reinforcement learning further enhancing capabilities. Use when you need LLMs to perform tasks requiring…
Idioma del texto original: inglés
Learn to encode LLM-generated responses rather than raw inputs by training special tokens and projection layers while keeping the backbone frozen. Bridges the input-output gap to transfer LLM capabilities like reasoning directly into embedding space.
Idioma del texto original: inglés
Accelerate image generation by generating multiple patches in parallel instead of sequentially. Uses locality-aware ordering that generates spatially-close tokens while keeping concurrent tokens far apart, reducing steps from 256 to 20 (12× speedup) while…
Idioma del texto original: inglés
Mechanistic Interpretability (MI) has emerged as a vital approach to demystify the opaque decision-making of Large Language Models (LLMs). However, existing reviews primarily treat MI as an observational science, summarizing analytical insights while lacking…
Idioma del texto original: inglés
Reduce self-attention complexity from O(N²) to O(N log N) through hierarchical token selection and enrichment. Perform hierarchical Top-K selection progressively adopting sparse Top-K at each level. Implement sparse index transpose algorithm avoiding dense…
Idioma del texto original: inglés
Train LLMs for long-context reasoning using KeyChain synthesis: convert short multi-hop QA into long-context tasks by embedding UUID chains in distractor documents, enabling 16K→128K generalization.
Idioma del texto original: inglés
Navigate Long CoT Degradation phenomenon when training small models on extended reasoning, understanding recovery dynamics and implementing strategies to maintain performance.
Idioma del texto original: inglés
Implements Controllable Memory Usage from arXiv:2601.05107
Idioma del texto original: inglés
Generate animations longer than 500 frames with consistent coloring. Uses dynamic global-local memory to compress long-term history and intelligently fuse global color features with local generation, enabling 49% quality improvement over previous methods.
Idioma del texto original: inglés
Implement techniques from LongCat-Flash-Thinking-2601 Technical Report. LongCat-Flash-Thinking-2601 achieves state-of-the-art performance among open-source models on a wide range of agentic benchmarks, including agentic search, agentic tool use, and…
Idioma del texto original: inglés
Train reward models to evaluate long-context responses by introducing faithfulness as a criterion and using consistency-based alignment to maintain judgment-explanation coherence.
Idioma del texto original: inglés
Training-free method extending diffusion LLMs to 6x context length using NTK-based RoPE scaling, exploiting bidirectional attention stability.
Idioma del texto original: inglés
Generate consistent ultra-long videos (up to one minute) using depth and trajectory controls with autoregressive segment generation and unified noise initialization.
Idioma del texto original: inglés
Train vision-language models on hour-long videos using a two-stage pipeline combining supervised fine-tuning with GRPO, reaching 71% accuracy on VideoMME while supporting 8,192 frames through efficient multi-modal sequence parallelism.
Idioma del texto original: inglés
Evict unnecessary KV cache entries using trainable lookahead tokens and LoRA modules that predict token importance without drafting responses. Achieve 14.5× speedup over draft-based methods with 2% eviction overhead.
Idioma del texto original: inglés
Enable budget-conditioned reasoning by repeatedly applying a shared transformer block stack with trajectory-based conditioning on time and step size. Train via shortcut-consistency loss to align shorter and full-length trajectories, enabling variable-depth…
Idioma del texto original: inglés
Improve LLM tool-use through automated, closed-loop data curation combining capability probing, error correction, and targeted data expansion—enabling self-refining training pipelines that surpass larger models without expensive APIs.
Idioma del texto original: inglés
Maximize parallelism in diffusion language models by intelligently exploring token filling orders. Spawn multiple candidate branches predicting high-confidence positions, evaluate all branches in one forward pass, and select the branch enabling most future…
Idioma del texto original: inglés
Preserve reasoning diversity in RL by protecting low-probability tokens that represent novel thinking paths. Trigger: maintain exploration capability during RL training while avoiding entropy collapse on reasoning tasks.
Idioma del texto original: inglés
Filter training samples by response length to identify high-confidence correct solutions and complex problems, improving sample efficiency in LLM reasoning RL without additional metrics.
Idioma del texto original: inglés
Improves instruction-following by using differentiated reward mechanisms that recognize logical dependencies (sequential, conditional, parallel) in instructions, enabling better model reasoning about task structure.
Idioma del texto original: inglés
Multimodal system for joint audio-video understanding and generation, enhancing agent perception and communication in environments with both audio and visual information.
Idioma del texto original: inglés
Build generalist agents for 3D open worlds using vision-language models with human-like interaction patterns and adaptive reasoning—achieving zero-shot transfer across games without fine-tuning through pixel-level perception and high-frequency action…
Idioma del texto original: inglés
Generate videos autoregressively by extending LLM architectures to spatiotemporal data. MM-RoPE balances frequency spectra across temporal and spatial dimensions, while Autoregressive Diffusion Forcing enables efficient parallel decoding. Lumos-1 (0.5B-3B…
Idioma del texto original: inglés
Fix length bias in sequence-level policy optimization by scaling each sequence's loss by its token count, eliminating gradient imbalances that cause models to shorten responses during training.
Idioma del texto original: inglés
Classify attention heads into retrieval (full attention) and sparse (token-selected) roles using HardKuma distribution for differentiable discrete optimization. Sparse heads reuse KV pairs from retrieval heads, reducing cache by 90% while maintaining quality…
Idioma del texto original: inglés
Enable models to stop generating when confident through lightweight hidden-state probes with distributional guarantees. LYNX achieves cross-domain transferability without retraining—ideal when you need confidence-controlled reasoning efficiency.
Idioma del texto original: inglés
Design and evaluate multimodal agents for tool use with M3-Bench: assess three interconnected dimensions (multi-modal grounding, multi-hop causality, multi-threaded parallelism) using similarity-bucketed Hungarian alignment for transparent tool call…
Idioma del texto original: inglés
Generates molecules meeting precise numeric property constraints across multiple dimensions through two-stage multi-agent framework with fragment-level edits and Group Relative Policy Optimization, improving validity and property satisfaction.
Idioma del texto original: inglés
Decomposes document processing into specialized agents (planning, execution, judgment, answer) with agent-wise adaptive test-time scaling. Achieves 9.9-11.5% performance gain with smaller models while maintaining reasoning.
Idioma del texto original: inglés
Build reasoning capabilities through pure RL without distilled traces, achieving 50% AIME accuracy improvement via scalable asynchronous training with novel reward shaping for multilingual consistency.
Idioma del texto original: inglés
Memory management architecture using multi-graph representations for autonomous agents, enabling efficient knowledge organization, contextual retrieval, and dynamic memory expansion to support complex agent decision-making.
Idioma del texto original: inglés
Scale GUI agents to real-world complexity via extended action space (user interaction, tool calls) and device-cloud collaboration. Online RL supports 500+ parallel environments with asynchronous handling; local agent monitors trajectory alignment and handoffs…
Idioma del texto original: inglés
Improve LLM training efficiency through manifold-based optimization that projects momentum onto tangent spaces and constrains updates on rotational Oblique manifolds. Achieves 1.75× faster convergence than Muon with reduced memory.
Idioma del texto original: inglés
Dynamically reweight advantage functions based on trajectory certainty to improve policy optimization in foundation models. Addresses advantage reversion and mirror problems by mixing standardized and mean-normalized advantage formulations. Enables more…
Idioma del texto original: inglés
Adapt DPO temperature parameters per sample using reward model margins, amplifying learning signals for hard preference pairs while dampening easy ones.
Idioma del texto original: inglés