논문

“Hy-Nomas is a video world model that uses causal attention and structured conditioning to generate interactive, multi-turn vide…”

나온 때 2026-10-26

가져온 때 2026-10-28 23:01 UTC · ed5f2c71ab05e085

뜯은 자리 — 초록

원제 — S-Space: Exploring Spatial Workspace in Multimodal Language Models

“Multimodal language models (MLLMs) exhibit promising spatial reasoning capabilities, yet the internal mechanisms supporting the…”

나온 때 2026-10-09

가져온 때 2026-10-12 23:01 UTC · dd7bd54bf7643623

뜯은 자리 — 초록

원제 — ProgramGym: Scaling Training Environments for Long-Horizon Program Reproduction Agents

“Coding agents are increasingly used to build programs from scratch as software engineers do rather than to resolve issues withi…”

대상 주장 — Fine-tuning on trajectories with these constructed environments raises the performance of Qwen3.8-27B from 34.87% to 42.84%.

나온 때 2026-10-08

가져온 때 2026-10-12 23:01 UTC · 221fc2bb32cab742

뜯은 자리 — 초록

원제 — EmbodiedCLBench: Evaluating Continual Learning for Self-Evolving Embodied Agents

“Continual learning requires models to accumulate past experience and transfer it to new tasks. Despite substantial research on …”

나온 때 2026-10-08

가져온 때 2026-10-12 23:01 UTC · 1725bcbca9c55585

뜯은 자리 — 초록

원제 — Continually Post-training Small Models via Synthetic Self-distillation

“Continual post-training faces a dual challenge of acquisition and retention: models must learn new capabilities while preservin…”

대상 주장 — Evaluations spanning 4 acquisition datasets (including a mixture) and 14 retention benchmarks show SynthSelf consistently outperforms 6 methods (including off- and on-policy distillation)…

나온 때 2026-10-08

가져온 때 2026-10-12 23:01 UTC · ecac050336939f35

뜯은 자리 — 초록

원제 — Work2Skill: Capturing Human GUI Experience for Reusable Agent Skill Induction

“Computer-use agents can perform increasingly complex GUI tasks, yet unfamiliar software and workflows requiring human domain kn…”

대상 주장 — When the same four execution models are equipped with skills induced by each method under shared execution settings, Work2Skill achieves the highest scores and success rates.

나온 때 2026-10-08

가져온 때 2026-10-12 23:01 UTC · 0bf731e1245f4493

뜯은 자리 — 초록

원제 — SepGen: Multi-Stem Audio-Video Separation and Generation in a Single Model

“A 4D audio-visual scene comprises a video, the dynamic geometry it depicts, and the sound sources that populate it, each with i…”

나온 때 2026-10-07

가져온 때 2026-10-12 23:01 UTC · f8acf62ebace4a63

뜯은 자리 — 초록

원제 — DiffusionART: Efficient, Stable and Scalable Reinforcement of Diffusion Models

“Online reinforcement learning (RL) fine-tunes diffusion models with black-box rewards, but reverse-process methods such as Flow…”

대상 주장 — In particular, on HunyuanVideo-13B with equally weighted visual and motion quality rewards, DiffusionART reaches the late-stage mean reward levels of DiffusionNFT and Flow-GRPO-Fast with …

나온 때 2026-10-07

가져온 때 2026-10-12 23:01 UTC · 789a779d68f1191e

뜯은 자리 — 초록

원제 — Reflex: Visually Grounded Reactive Humanoid Control

“Humans effortlessly react to the world; they adeptly return a tennis shot, catch a falling cup, step aside from a passing cart,…”

대상 주장 — In simulation, Reflex achieves 85.7% catch success from egocentric RGB-D, compared with 87.9% when given privileged box observations.

나온 때 2026-10-07

가져온 때 2026-10-12 23:01 UTC · a3843271a70bb133

뜯은 자리 — 초록

원제 — SoftwareGym: Scaling Computer-Use Agent Training with Diverse Synthesized Software Environments

“Computer-use agents (CUAs) are rapidly improving, yet synthesizing diverse and reliable software environments at scale remains …”

대상 주장 — Using SoftwareGym, we synthesize 182 applications across twelve categories and achieve a mean GUI task pass rate of 91.5%.

나온 때 2026-10-07

가져온 때 2026-10-12 23:01 UTC · 5a26c82bd2ed4195

뜯은 자리 — 초록

원제 — VisionWeave: Weaving Elastic Visual Representations as a Native Capability of MLLMs

“Multimodal large language models have become the dominant paradigm for visual understanding, but incur substantial costs by enc…”

대상 주장 — Based on Qwen3.8-27B, VisionWeave adaptively adjusts token savings to visual content, saving 43.0% tokens on average while retaining 98.9% native performance across eight benchmarks, vers…

나온 때 2026-10-05

가져온 때 2026-10-12 23:01 UTC · 57525ad9734267df

뜯은 자리 — 초록

원제 — SoloQ: Calibration-Free Quantization for Diffusion Language Models

“Diffusion large language models dLLMs) have emerged as a promising alternative to autoregressive language models through bidire…”

대상 주장 — Across full-sequence dLLMs (LLaDA and Dream) and block-diffusion dLLMs(Fast-dLLM v2 and Nemotron-Labs-Diffusion), SoloQ retains accuracy under 4-bit quantization and outperforms calibrati…

나온 때 2026-10-04

가져온 때 2026-10-12 23:01 UTC · 2ba5dc712e71f81e

뜯은 자리 — 초록

원제 — Here the World in Stereo: Learning Dynamic Spatial Correspondence for Immersive Joint Video-Audio Generation

“Recent joint video-audio generation models have achieved strong semantic correspondence and temporal synchronization. However, …”

나온 때 2026-09-29

가져온 때 2026-10-12 23:01 UTC · 7768b360a8a3a4b5

뜯은 자리 — 초록

원제 — DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training

“We present DreamTrue, a multi-view, cross-embodiment robot world model for action-faithful and physically plausible video predi…”

대상 주장 — On AgiBot, DreamTrue attains state-of-the-art action following, while reducing the human-assessed interaction defect rate from from 48.12% to 6.25%.

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 074b3efb8c207b56

뜯은 자리 — 초록

원제 — Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction

“Egocentric world models predict first-person observations conditioned on an agent's actions, but most focus on a single agent. …”

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 984783c99f6c9c63

뜯은 자리 — 초록

원제 — OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D Cinemagraphs

“Recent 3D world models generate photorealistic, explorable scenes that remain frozen in time. OuroWorld is a mask-free framewor…”

대상 주장 — On 39 reconstructed and generated scenes, OuroWorld outperforms all baselines and wins 70.8%-99.0% of user-study comparisons.

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 2df83b14d4e56c3d

뜯은 자리 — 초록

원제 — Foundations of Large Language Models

“This is a book about large language models. As indicated by the title, it primarily focuses on foundational concepts rather tha…”

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 5a28cbe0f0dba767

뜯은 자리 — 초록

원제 — Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks

“Learning to act in unfamiliar environments requires agents to infer how the world works and revise that understanding as new ev…”

대상 주장 — On ARC-AGI-3, the single-model agent clears every level of all 25 public games, achieves a mean Relative Human Action Efficiency (RHAE) of 100.0, and uses 44% of the human action count.

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 2ec664cc3165cff7

뜯은 자리 — 초록

원제 — Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching

“Dense correspondence matching has historically been bounded by simplifying spatio-temporal priors, such as smooth motion and ri…”

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 928c5916c8745d83

뜯은 자리 — 초록

원제 — MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement

“Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement. Th…”

대상 주장 — We scale RL compute along three dimensions: (1) larger batches and higher throughput, with an asynchronous training that consumes 1,568 samples and 2.7-3.7B tokens per step at context len…

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 1b5db0f1008c3cfe

뜯은 자리 — 초록

원제 — TokenRouter: Efficient Serving System for Token-Level LLM Routing

“Large language model (LLM) routing distributes inference work across different models, advancing the cost-quality Pareto fronti…”

대상 주장 — Across diverse routing algorithms, workloads, and model pairs, TokenRouter achieves 2.01-64.15x higher decoding throughput than existing systems, substantially advancing the serving effic…

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · e1c4e3f59be4b797

뜯은 자리 — 초록

원제 — REMORY: Learning Residual Memory for Context Compaction

“Long-horizon agents compact their history to continue within a finite context window, but a textual summary alone may not suppo…”

대상 주장 — On SummHay, REMORY improves source attribution at nearly unchanged insight coverage and approaches the full-context joint score using only 5.2% of the input positions.

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · d6956d1b7ab9450f

뜯은 자리 — 초록

원제 — Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments?

“Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability i…”

대상 주장 — (2) Human agent gap: Top-performing human players achieve higher peak scores than the evaluated agents.

나온 때 2026-10-07

가져온 때 2026-10-09 23:01 UTC · 2a9358bd3d19b21d

뜯은 자리 — 초록

원제 — U-Space: Uncovering When and Why Uncertainty Arises in Language Models

“Large language models are informing decisions with ever-higher stakes. As the consequences of their errors grow, a central ques…”

나온 때 2026-10-05

가져온 때 2026-10-09 23:01 UTC · b917d3ea1ea8f81a

뜯은 자리 — 초록

원제 — In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks

“We study robotic in-context learning (ICL), an emerging paradigm that enables robots to infer and execute tasks from visual dem…”

나온 때 2026-09-28

가져온 때 2026-10-09 23:01 UTC · 8e9c091792f18ffc

뜯은 자리 — 초록

원제 — FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders

“Representation autoencoders (RAEs) reuse features from a pretrained visual encoder as reconstruction and diffusion latents, int…”

대상 주장 — On ImageNet-256 with DINOv3-L, a single FuseReg decoder reconstructs from full, sparse, and single-layer fusions without retraining, achieving higher PSNR than decoders specialized to fix…

나온 때 2026-09-24

가져온 때 2026-09-28 05:05 UTC · 6861a4d455b6989e

뜯은 자리 — 초록

원제 — InternW0-Δ: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data

“World Action Models (WAMs) jointly model visual dynamics and action generation for generalist robot manipulation. A central cha…”

대상 주장 — We introduce InternW0-Δ, a unified WAM pretrained on a heterogeneous corpus that outperforms prior methods across simulation benchmarks and real-robot platforms.

나온 때 2026-09-24

가져온 때 2026-09-28 05:05 UTC · fce8fc3803ad1364

뜯은 자리 — 초록

원제 — Block Sparse Attention with Log-Linear Complexity

“Scaling language models to long contexts is limited by the quadratic cost of self-attention. Block sparse attention offers an e…”

나온 때 2026-09-24

가져온 때 2026-09-28 05:05 UTC · e8ca90729f4aeb76

뜯은 자리 — 초록

원제 — Qwen-Planner-Agent: A Closed-Loop AI-for-AI Framework for Real-World Mobile Planner Agents

“The rapid progression of large language models is extending AI from passive content generation into the active workflows of eng…”

나온 때 2026-09-23

가져온 때 2026-09-25 23:01 UTC · 4f969e088a867b41

뜯은 자리 — 초록

원제 — Coding Agents for Generalized Task and Motion Planning Problems

“Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discre…”

대상 주장 — Across all program synthesis methods, we evaluate 980 generated programs on 100 held-out instances each, 98,000 evaluation episodes in total.

나온 때 2026-09-23

가져온 때 2026-09-25 23:01 UTC · 8f9867a434213048

뜯은 자리 — 초록