ADu2021/skillXiv - Page 15

latcoder-layout-aware-code-generation

Convert webpage designs to code via Layout-as-Thought reasoning, detecting layout structure and generating HTML/CSS for spatial blocks.

latent-chain-of-thought

Improve reasoning quality by decoupling reasoning from verbalization. Planner generates deterministic latent trajectories while Decoder grounds them to text. Enables dynamic termination, better Pass@k scaling, and interpretable intermediate states.

latent-collaboration-multiagent

Enable LLM agents to collaborate entirely in latent space by sharing layer-wise KV cache representations instead of text, achieving 4× speedup and 71-84% fewer output tokens while maintaining 14.6% higher accuracy through lossless information preservation in…

latent-diffusion-self-supervised

Replace VAE in latent diffusion with frozen DINO self-supervised features plus lightweight residual processing, enabling faster training, few-step sampling, and clearer semantic structure.

latent-entropy-aware-decoding

Reduce hallucinations in multimodal reasoning by detecting high-entropy (uncertain) states and switching to continuous latent embeddings instead of discrete tokens. Use prior-guided visual anchoring during uncertain phases to maintain grounding.

latent-particle-world-models

Learn world models by decomposing scenes into latent particles with per-particle dynamics. Each particle represents an object with position, scale, and appearance. Learn distributed latent actions governing per-particle transitions, enabling multimodal video…

latent-sketchpad-visual-reasoning

Enables multimodal reasoning by interleaving visual sketches with text. MLLMs generate latent visual representations during reasoning, with sketch decoder converting them to human-interpretable images. Improves reasoning performance while maintaining…

latent-thoughts-tuning

Enable models to reason in continuous latent space via context-prediction fusion, combining hidden state context with vocabulary embeddings to maintain scaling across model sizes.

latte-flow-unified-multimodal

Unify image understanding and generation with layerwise timestep experts and residual attention reuse, achieving 6x faster inference than comparable unified models while maintaining competitive performance.

lattice-3d-generation

Semi-structured latent representation combining efficiency of VecSet with spatial structure guidance via voxel queries and rotary positional embeddings, enabling strong test-time scaling (6K to 30K tokens) and improved model scaling without sparse components.

lavit-visual-reasoning

Improves vision-language model distillation by aligning latent visual reasoning trajectories between teacher and student, enabling 3B parameter models to outperform larger open-source and proprietary systems with +16.9% gains on reasoning tasks.

layer-cake-contrastive-decoding

Improve factual accuracy in LLM generation through decoding-time layer-wise attention suppression. Selectively suppress attention to specific token types at their most influential transformer depths without training or model modifications. Use when you need…

leank-kv-cache-pruning

Learning-based approach to reduce key-value cache memory during inference through static channel-wise sparsity masks. Achieves 70% K cache and 16-18% V cache memory reduction with 1.3x attention speedup.

learnable-multipliers-lm-scaling

Adapt language model weight matrix scales automatically during training by adding learnable scalar and vector multipliers to each layer. Breaks the noise-weight-decay equilibrium that constrains scales based on hyperparameters rather than data, enabling…

learned-4bit-quantization

Compress LLM weights to 4-bit precision using learned per-row lookup tables that minimize output activation error, achieving better accuracy than fixed formats (int4, fp4, nf4) while maintaining inference speed.

learning-4d-reasoning

Enable VLMs to perform dynamic spatial reasoning (DSR) by extracting 4D priors from videos and using Geometry Selection Modules (GSM) for selective injection. Provides DSR-Train dataset (50K QA pairs) and benchmark with six reasoning types, balancing…

learning-beyond-teacher

Train student models to exceed teacher performance by extrapolating rewards during distillation using a scaling factor λ1. Enables unified students to surpass all individual domain teachers through controlled reward amplification.

learning-discover-test-time

Apply reinforcement learning at test time to enable language models to continue adapting on specific problems, achieving state-of-the-art results across mathematics, GPU optimization, algorithms, and biology. Use when you need models to discover…

learning-on-job-self-evolving-agent

Build autonomous agents that accumulate structured knowledge from task execution into hierarchical memory (strategic, procedural, tool) without human annotation, enabling knowledge transfer to unseen tasks.

learning-on-the-job-test-time-curricula

Enable models to autonomously improve on target tasks during inference via test-time curricula (TTC-RL). Automatically select task-relevant training examples and apply RL to continue learning. Achieve 1.8x improvement on AIME25 math benchmarks and 2.1x on…

learning-reason-factuality

Multi-dimensional reward function combining factual precision, response detail, and answer relevance for online RL. Reduces hallucinations 23.1% while maintaining helpfulness and detail.

learning-scientific-taste-with-rl

Learn to predict and generate high-impact research ideas by training models on community feedback signals. Apply reinforcement learning to align research generation with citation-based indicators of scientific impact.

learning-user-preferences-collaboration

Enable agents to adapt to individual user preferences across extended interactions via memory-equipped architectures. Use learning signals to improve memory updates and enable long-term collaborative relationships with human users.

learning-visual-priors-llm-pretraining

Decompose visual priors into perception and reasoning components, each optimized by distinct data types: reasoning from code/math corpora, perception from diverse modality-rich sources. Use to construct efficient VLM pretraining pipelines balancing multimodal…

lemas-multilingual-speech-dataset

Release 150K-hour multilingual speech corpus with word-level timestamps plus non-autoregressive flow-matching TTS and autoregressive speech editing models. Enable diverse speech generation and editing tasks across 10 languages.

length-incentivized-exploration

Overcome the shallow exploration trap by explicitly rewarding longer reasoning sequences when models fail to solve problems. Use length-incentivized exploration to enable deeper chain-of-thought reasoning, achieving better test-time scaling and improved…

less-is-more-recursive-reasoning-tiny-networks

Achieve complex reasoning with minimal parameters using latent recursion in 2-layer networks. A 7M-parameter Tiny Recursive Model (TRM) solves Sudoku (87% accuracy), mazes (85%), and ARC-AGI with 0.01% the parameters of large LLMs via iterative latent…

less-is-more-sparse-attention

Improves inference efficiency through training-free sparse attention using global token selection patterns aggregated from local attention heads for unified cross-head ranking.

let-it-calm-annealed-decoding

Use dynamic temperature scheduling that explores early (high temperature) and exploits late (low temperature) during generation, improving sample efficiency in RL with verifiable rewards.

let-late-early-training

Accelerate LLM pretraining by leveraging small pretrained models as teachers, enabling large models to explicitly learn late-layer knowledge in early layers via alignment loss, achieving 1.6x convergence speedup and 5% downstream improvement even with 10x…

liberty-a-causal-framework-for-benchmarking

Concept-based explanations quantify how high-level concepts (e.g., gender or experience) influence model behavior, which is crucial for decision-makers in high-stakes domains. Recent work evaluates the faithfulness of such explanations by comparing them to…

light-if-preview-checking

Multi-stage training approach using entropy-preserving SFT and token-wise entropy-adaptive RL to improve instruction adherence. Combines data curation with reward-guided reasoning, outperforming larger models on IFEval.

lightweight-memory-augmented

Reduce inference tokens by up to 38× and API calls by 30× through three-stage memory system: sensory compression, short-term consolidation, and offline long-term storage.

listener-rewarded-vlm-preferences

Aligns reasoning traces with final decisions in preference models using an independent frozen VLM as listener. Achieves 67.4% accuracy on ImageReward by enforcing consistency between explanations and choices. Use when training reward models for image…

lite-attention-temporal-sparse

Accelerate video diffusion generation by exploiting temporal attention sparsity—skip redundant attention tiles across denoising steps using persistent skip masks, achieving 40% speedup with quality retention.

livemcpbench-agent-tool-navigation

Benchmark framework for evaluating LLM agents navigating large-scale Model Context Protocol ecosystems with multi-tool composition across 95 daily tasks.

livr-visual-reasoning

Enable richer visual reasoning in multimodal models via learnable latent tokens trained with visual bottlenecking. Introduces special tokens that implicitly learn task-relevant visual abstractions without explicit supervision, through attention masking…

lk-losses-speculative-decoding

LK Losses replace KL divergence with direct acceptance rate optimization for speculative decoding, achieving 8-10% improvements without computational overhead.

llada2-diffusion-scaling

Convert pre-trained autoregressive models into large-scale diffusion language models via Warmup-Stable-Decay training strategy. Progressively increase block size during warmup, perform stable diffusion training, then decay to smaller blocks for inference.…

llds-grpo-collapse

Identifies Lazy Likelihood Displacement (LLD) as GRPO failure mechanism in tool-integrated LLMs and proposes lightweight LLDS regularization that penalizes only tokens responsible for likelihood decrease, achieving up to 45.2% performance gains.