cập nhật 36 phút trước

⏱ Độ trễ thật của tab này: trung bình 6.8 giờ sau khi CVE được công bố (n = 5264, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.

20
Hy-Nomas: Towards Free-form Interactive World Generation
📄 Paper HF Daily Papers 24 ngày trước

Interactive world generation requires models to follow successive commands while preserving relevant history and action consequences. We present Hy-Nomas, a video world model with an 8B diffusion backbone that supports free-form text instructions and continuous camera control acr…

🅲 hạng C
20
Look Only When Needed: Certificate-Guided Visual Routing for Efficient Diffusion VLMs
📄 Paper HF Daily Papers 24 ngày trước

Diffusion vision-language models (DVLMs) process their entire input at every denoising step. Since visual tokens make up 89-97% of this input on the benchmarks we study, most of the decoding computation is spent repeatedly processing the same image. Visual token pruning methods p…

🅲 hạng C
20
FrameForge: Harnessing Priors of Video Foundation Models for Unified Image Generation
📄 Paper HF Daily Papers 24 ngày trước

Unified conditional image generation requires a single model to preserve different aspects of a visual input while producing a target image that follows textual guidance. Pretrained video models provide a promising foundation by encoding visual transformations across time, yet ex…

🅲 hạng C
20
Turns Matter: Advancing Streaming Video RL via Turn-Level Credit
📄 Paper HF Daily Papers 16 ngày trước

Streaming video understanding requires models to repeatedly decide whether to respond or continue observing. Recently, Group Relative Policy Optimization (GRPO) has been applied to this setting, with existing approaches emphasizing reward design while retaining a single trajector…

🅲 hạng C
20
Can Discrete Tokens Serve as Semantically Lossless Representations of Visual Signals?
📄 Paper HF Daily Papers 24 ngày trước

Discrete visual tokens represent vision in a discrete form analogous to language, offering a native interface for multimodal modeling under a shared next-token prediction objective. However, quantization error is widely assumed to cause unavoidable semantic loss, fundamentally li…

🅲 hạng C
20
CUA-Master: Scaling Computer-use Agents through Open-World Recursive Improvements
📄 Paper HF Daily Papers 17 ngày trước

Despite recent advances in CUAs, their development remains constrained by the limited diversity, controllability, and reliability of interaction data. Moreover, conventional optimization pipelines lack iterative feedback, while existing recursive improvement approaches often rely…

🅲 hạng C
20
Route Anything: A Study of Model and Policy Routing for General-Purpose AI Systems
📄 Paper HF Daily Papers 17 ngày trước

Artificial general intelligence (AGI) requires a method that can understand, generate, reason, and act across diverse domains and modalities. Rather than relying on a single model to master every capability, a promising path toward AGI now is to orchestrate a heterogeneous ecosys…

🅲 hạng C
20
Learning Heterogeneous Tactile Representation for General Dexterous Manipulation
📄 Paper HF Daily Papers 11 ngày trước

Tactile sensing is an indispensable modality for contact-rich dexterous manipulation, a capability that hints at the ultimate frontiers of embodied intelligence. However, commercially available tactile sensors are still at a relatively early stage of development, with each type o…

🅲 hạng C
20
Proving and Improving the Platonic Representation Hypothesis with JEPAs
📄 Paper HF Daily Papers 22 ngày trước

Whether models trained on different modalities converge to the same representation of the world is central to AI. The Platonic Representation Hypothesis (PRH) conjectures that they do as they scale, but without a proof it offers neither a guarantee nor a recipe. We turn the hypot…

🅲 hạng C
20
Work Agents: Systems, Capabilities, and Optimization
📄 Paper HF Daily Papers 25 ngày trước

As foundation models evolve from information generation to digital work execution, agents are increasingly capable of completing complex tasks across tools, applications, and workflows. However, existing research lacks a unified perspective on how agents reliably accomplish deleg…

🅲 hạng C
20
Think Less, Locally: Local Preference Learning for Reasoning Compression
📄 Paper HF Daily Papers 17 ngày trước

Large reasoning models achieve strong performance on challenging tasks, but their long chain-of-thought often incurs substantial inference costs by continuing to deliberate after sufficient information has been obtained. Existing reasoning compression methods primarily operate on…

🅲 hạng C
20
From Pixels to Actions - Towards Native Vision-Language-Action Policies
📄 Paper HF Daily Papers 15 ngày trước

Vision-language-action models are typically built on a simple premise: perception is first compressed by visual encoders, and action reasoning begins only afterwards. Yet manipulation often hinges on precisely the visual signals most vulnerable to this abstraction, e.g., local ge…

🅲 hạng C
20
SplatAnything: Gaussian Spacing Normalization for Feed-forward 3DGS across Arbitrary Scenes
📄 Paper HF Daily Papers 14 ngày trước

We identify a key limitation of existing feed-forward 3DGS methods: Gaussian attributes are typically predicted in a radial-distance-normalized space that canonicalizes a global radial statistic of the reconstructed geometry, while the spacing between neighboring Gaussian centers…

🅲 hạng C
20
CodeCap-Bench: Can Code Serve as a Good Video Caption?
📄 Paper HF Daily Papers 24 ngày trước

Video captioning aims to preserve visual information through compact descriptions, predominantly using natural language. However, precisely expressing fine-grained spatial structures and temporal dynamics in natural language often leads to lengthy and potentially ambiguous descri…

🅲 hạng C
20
ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning
📄 Paper HF Daily Papers 24 ngày trước

Vision-Language-Action (VLA) models have made substantial progress in robotic manipulation, but their generalization remains constrained by the diversity and coverage of action-labeled robot demonstrations. Human videos provide much broader behavioral and environmental experience…

🅲 hạng C
20
DreamX: Video Diffusion Model is a Universal Latent Action Learner
📄 Paper HF Daily Papers 17 ngày trước

Learning a universal action representation from video requires modeling dynamics that are consistent across different executions of the same action yet invariant to changes in embodiment and appearance. Existing latent action models infer actions from visual transitions, but thei…

🅲 hạng C
20
Random Is Almost Enough? Stable and Low-Cost Benchmark Compression via BestRand
📄 Paper HF Daily Papers 20 ngày trước

Benchmark compression reduces the often prohibitive time and cost of model evaluation by evaluating models on a small benchmark subset that closely matches full-benchmark results. To quantify estimation error, we derive an accurate analytical approximation to the error distributi…

🅲 hạng C
20
WorldFoundry: Towards Unified Infrastructure for World Intelligence
📄 Paper HF Daily Papers 12 ngày trước

Visual generative models increasingly blur the line between synthesizing observations and simulating environments. Yet visual plausibility alone does not constitute successful world modeling: compelling outputs may lack persistent state, physical dynamics, or causally consistent …

🅲 hạng C
20
Sage-Taiyi: A Metacognitive Architecture for Scientific Discovery
📄 Paper HF Daily Papers 11 ngày trước

We introduce Sage-Taiyi, a metacognition-inspired architecture that uses accumulated scientific understanding to guide what to investigate next and how to investigate it. Through explicit monitoring and control, the system revises its understanding of evidence, assumptions, and u…

🅲 hạng C
20
Spatial Audio Generation and Editing with Executable Plans and Self-Distilling Transfusion
📄 Paper HF Daily Papers 26 ngày trước

Spatial audio authoring coordinates source decisions with soundfield synthesis. For editing, plans specify what to change and retain, while reference audio supplies further acoustic detail. We present AMBIT, a unified framework for native First-Order Ambisonics (FOA) generation a…

🅲 hạng C
20
Rollback: Experience Backtracking in World Models for Policy Improvement
📄 Paper HF Daily Papers 25 ngày trước

Action-conditioned world models offer a scalable alternative to costly real-world interaction for improving robot policies. However, existing world-model-based methods largely treat world models as forward-only surrogates of the physical environment, leaving untapped a capability…

🅲 hạng C
20
L3G-World: Real-Time Infinite World Generation with Latent 3D Gaussian Memory
📄 Paper HF Daily Papers 22 ngày trước

Interactive world models generate visual content conditioned on user actions, yet maintaining spatial consistency over long-horizon exploration in real time remains challenging. Existing approaches rely on either temporal context or explicit 3D memory, both of which scale poorly …

🅲 hạng C
20
Factorized MLLM Embeddings for Composed Image Retrieval
📄 Paper HF Daily Papers 17 ngày trước

Composed Image Retrieval (CIR) aims to retrieve a target image given a reference image and a modification text. Although recent approaches leveraging Multimodal Large Language Models (MLLMs) have shown significant progress, we argue that the MLLM's compositional understanding rem…

🅲 hạng C
20
FrontierPhysics: Evaluating Agents for End-to-End Physics Research
📄 Paper HF Daily Papers 16 ngày trước

AI agents are playing an increasingly important role in physics research. However, existing physics benchmarks only evaluate fixed questions, isolated sub-tasks, or paper reproduction with given solutions, and rarely check research quality alongside correctness. We introduce Fron…

🅲 hạng C
20
WeLM-Image: A Unified Image Generation Model towards Visual Intelligence
📄 Paper HF Daily Papers 15 ngày trước

We present WeLM-Image, a unified image foundation model that integrates multimodal understanding into text-to-image generation, instruction-guided editing, and multi-reference synthesis. Designed for single-GPU deployment in WeChat's large-scale consumer applications, the model p…

🅲 hạng C
20
RSIHub: A Composable Framework for Agent-Evolution Methods
📄 Paper HF Daily Papers 28 ngày trước

Agent-evolution methods use execution feedback to improve prompts, skills, tools, and orchestration, but their implementations often rebuild the same experimental machinery. We present RSIHub, an open-source framework for building, comparing, and extending these methods on shared…

🅲 hạng C
20
Sensorimotor Alignment: A Platonic Proxy for Embodied Policy Success
📄 Paper HF Daily Papers 28 ngày trước

Robot foundation models are typically evaluated through costly closed-loop rollouts, motivating low-cost offline proxies for model capability. The Platonic Representation Hypothesis (PRH) suggests that as representations improve, modalities such as vision and language converge to…

🅲 hạng C
20
Read, Allocate, Route: Role-Coupled Tokenization for Pixel-Space Generative Transformers
📄 Paper HF Daily Papers 22 ngày trước

Pixel-space generative Transformers typically use one token per image patch, so capturing finer detail requires more tokens and makes global attention increasingly expensive. We observe that finer detail does not necessarily require more global communication: low-frequency struct…

🅲 hạng C
20
Apriel Voice 27B Preview: An Open Model for Speech, Reasoning, and Tool Use
📄 Paper HF Daily Papers 22 ngày trước

A capable voice assistant must do more than recognize words; it must understand requests, reason over context, sustain conversations, and act through tools. We present Apriel Voice 27B Preview, an open speech-language model that unifies these capabilities within a shared reasonin…

🅲 hạng C
20
Democratizing Robot-Use Agents: A Unified Harness with Open Models
📄 Paper HF Daily Papers 19 ngày trước

Robot-use Agents (RUAs) have shown promising zero-shot manipulation capabilities, but reliance on frontier, proprietary models limits research into further training. To democratize RUA research, we introduce uni-robot-agent, a unified harness for competitive open-weight robot use…

🅲 hạng C
20
Halley: A Predictive Control Plane for Optical AI Data Center Networks
📄 Paper HF Daily Papers 16 ngày trước

A frontier AI model is trained in an AI data center, on tens of thousands of accelerators joined by a dedicated network, and every training step ends with all of them exchanging data at once. Between racks, that network is optical, and each of its four layers, from lasers to swit…

🅲 hạng C
20
TailSteer: Trajectory-Tail Steering for Drift-Resistant Multi-turn Image Editing
📄 Paper HF Daily Papers 16 ngày trước

Strong single-turn editing performance does not guarantee reliable multi-turn interaction: as edited outputs are repeatedly reused, small unintended changes can accumulate into severe visual drift. This long-term drift is often mitigated through training on multi-turn data or cro…

🅲 hạng C
20
P^2-MOPD: Accelerating Multi-Teacher On-Policy Distillation via Parameter Priors
📄 Paper HF Daily Papers 17 ngày trước

Multi-teacher on-policy distillation consolidates expert capabilities by supervising student-generated trajectories, whereas task arithmetic offers a training-free alternative through linear parameter combinations. We view the linear combination of student and expert parameters a…

🅲 hạng C
20
G-STAR: Group-Scaled Token Capacity for Deeply Compressed Autoregressive Generation
📄 Paper HF Daily Papers 12 ngày trước

Discrete autoregressive models offer a scalable framework for image generation. However, they face a fundamental trade-off: long token sequences preserve representation quality but slow generation, while aggressive spatial compression reduces representation capacity. Based on pro…

🅲 hạng C
20
Open Vision Banana: A Fully Open Recipe for Generative Visual Perception
📄 Paper HF Daily Papers 9 ngày trước

Recent work suggests that pretrained image generators encode real-world understanding beyond image synthesis. Vision Banana exploits this by framing vision tasks as conditional image editing. However, its data, weights, code, and implementation details are not public. We present …

🅲 hạng C