cập nhật 22 phút trước

⏱ Độ trễ thật của tab này: trung bình 6.6 giờ sau khi CVE được công bố (n = 5196, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.

20
Alloy: A Multi-Material Physics Engine for Robot Learning
📄 Paper HF Daily Papers 9 ngày trước

Robots handle much more than rigid objects: towels, cables, water, sand. The simulators used to train robots at scale are strongest on rigid bodies, and adding another material usually means a second engine and hand-written glue between the two. We present Alloy, an open-source p…

🅲 hạng C
0
AdaPRA: Process Reward Agents with Adaptive Rubrics for Data Science
📄 Paper HF Daily Papers 29 ngày trước

Data science agents can automate end-to-end machine learning workflows, yet they remain far from competitive on human leaderboards. Agents make process errors such as insufficient exploration, unsound validation, coding bugs, and poor use of a fixed budget (re-running the same ex…

🅲 hạng C
0
A Broader Look into MLLM's Retrieval Heads
📄 Paper HF Daily Papers 29 ngày trước

Locating relevant evidence in long inputs is a core capability of Multimodal Large Language Models. Prior studies attribute this capability to retrieval heads, a sparse set of attention heads that transfer information from the context to the output. However, these analyses are co…

🅲 hạng C
0
CEPO: A General Framework for Improving Prompt Optimization in Multi-Agent Systems
📄 Paper HF Daily Papers 26 ngày trước

Multi-agent systems (MAS) require both effective coordination and specialized task execution. Existing prompt optimization approaches typically couple these functions within agent-specific prompts. We introduce Coordinator-Executor Prompt Optimization (CEPO), a general framework …

🅲 hạng C
0
OSWAM: An Effective World Action Model with One-Step Action Generation
📄 Paper HF Daily Papers 26 ngày trước

Direct-policy World Action Models (WAM) have demonstrated the benefits of decoupling action generation from explicit future visual prediction for efficient inference in autonomous driving and embodied control. However, they still require tens of denoising steps to achieve competi…

🅲 hạng C
0
SOMA: A Benchmark for Fine-Grained Egocentric Behavior and Body-Motion Understanding
📄 Paper HF Daily Papers 26 ngày trước

Perceiving human body motion is critical for multimodal systems that understand and assist people. Yet current evaluations of Multimodal Large Language Models (MLLMs) largely focus on objects, scenes, and coarse actions, leaving their ability to reason about the body itself poorl…

🅲 hạng C
0
GEPA-Curator: Curated Learning for Self Improving Agents
📄 Paper HF Daily Papers 25 ngày trước

Large Language Models (LLMs) are increasingly used in multi-turn agentic workloads that require sustained interaction with complex environments. Existing reflective prompt optimizers such as GEPA learn from failed rollouts, but agentic failures can also come from prompt weaknesse…

🅲 hạng C
0
Multi-Agent Relay Thinking
📄 Paper HF Daily Papers 25 ngày trước

Existing multi-agent collaboration (MAC) methods have primarily explored a group thinking collaboration topology, where agents independently generate solutions and aggregate their outputs. In this work, we investigate an alternative relay thinking topology, where reasoning owners…

🅲 hạng C
0
Advancing SVD-based LLM Compression via Layer-Wise Error Model Search
📄 Paper HF Daily Papers 29 ngày trước

Low-rank SVD-based compression offers a powerful strategy to reduce the computational costs of LLMs. However, existing methods face two key limitations: (i) global rank allocation, where uncalibrated error proxies fail to capture complex error propagation, and (ii) decomposition …

🅲 hạng C
0
A Tutorial on Generative Models for Music Audio
📄 Paper HF Daily Papers 24 ngày trước

Music audio generation spans instrumental clips, full songs with vocals, edits of existing recordings, and systems that play alongside a musician in real time. This monograph explains the methods behind these systems for readers familiar with neural-network training who are new t…

🅲 hạng C
0
LOLPlay: Benchmarking Real-Time Visual Agents in League of Legends
📄 Paper HF Daily Papers 27 ngày trước

Visual agents need to coordinate reasoning and action while their environment evolves during inference. We introduce LOLPlay, a benchmark in native League of Legends with 36 champion-specific tasks spanning resource management, combat, cooperation, and objectives. Agents interact…

🅲 hạng C
0
Muon Makes the First Mixture-of-Experts Layer Harder to Balance
📄 Paper HF Daily Papers 25 ngày trước

Many open mixture-of-experts (MoE) language models train with Muon instead of AdamW. Most balance the load with a bias on each expert's routing score, adjusted after every step. Their recipes disagree on whether Muon should also train the router. In models with 55M and 108M activ…

🅲 hạng C
0
How I Aligned Your Model: Scaling Up Multilingual Safety Training with Templates
📄 Paper HF Daily Papers 23 ngày trước

Safety alignment for Large Language Models is heavily focused on English with limited multilingual representation, particularly damaging low-resource language performance. Related open training datasets are often limited in size or do not follow region-localized harmful topics to…

🅲 hạng C
0
Identically Trained Language Models Break on Different Jailbreaks
📄 Paper HF Daily Papers 27 ngày trước

Jailbreak benchmarks test one checkpoint and treat which jailbreaks break it as a property of its pretraining data and safety training. We ask whether another pretraining run of the same data, given the same safety training, would break on the same jailbreaks. For each of 25 pret…

🅲 hạng C
0
SynthFaces: Balanced Large Scale Human Dataset
📄 Paper HF Daily Papers 23 ngày trước

The pursuit of fair and equitable AI in human-centric tasks is critically hindered by the inherent biases and privacy risks of real-world datasets, which offer imbalanced representations of identity and often contain sensitive personal information. Consequently, state-of-the-art …

🅲 hạng C
0
Hard Sample-Conditioned Decorrelated Attention for Robust Face Recognition
📄 Paper HF Daily Papers 21 ngày trước

Multi-head attention in Vision Transformers (ViTs) allows independent attention patterns across heads, but without explicit regularization, heads may learn redundant patterns rather than complementary roles. Correlation-based penalties encourage head disentanglement, yet apply un…

🅲 hạng C
0
AgentBaton: Efficient Model Collaboration in Agentic Tasks
📄 Paper HF Daily Papers 27 ngày trước

Solving long-horizon agentic tasks requires language agents to interact with their environments over multiple steps, and reliable performance often comes from the expensive capable models available through cloud providers. Meanwhile, alternative efficient models with inferior per…

0
"PRISM" Is Not All You Need: Cross-Topic Name Collisions and Scientific Paper Retrieval
📄 Paper HF Daily Papers 20 ngày trước

Scientific names increasingly connect papers from different research topics, leaving readers to recover a paper's identity from context. We study this convergence in 564,861 AI-related arXiv records. Cross-topic name collision grows 4.33-fold from a pooled 2020-2022 reference to …

🅲 hạng C
0
PREEN: Exemplar-Based Color Style Transfer with Smartphone Editor Presets
📄 Paper HF Daily Papers 25 ngày trước

We present PREEN (Preset Regression via an Exemplar-based Editor Network), which casts color style transfer as regression onto the 29 parameters of a deterministic image editor. Given a content image and a style reference, PREEN predicts a preset, a vector of editor slider values…

🅲 hạng C
0
Fast and Exact Relative-Position Attention via Structure-Aware Kernels
📄 Paper HF Daily Papers 21 ngày trước

Relative-position attention mechanisms offer compelling alternatives to RoPE across various transformer architectures, delivering superior performance in domains such as audio. These mechanisms typically modify attention scores by adding content-position or bias-position terms. U…

🅲 hạng C
0
Learning to evolve by matching velocities
📄 Paper HF Daily Papers 21 ngày trước

Learning the continuous-time evolution of nonlinear dynamical systems relies on trajectory-based objectives, which require simulating the learned dynamics during training and backpropagating through the numerical scheme, at a cost that grows with the rollout horizon. We show that…

🅲 hạng C
0
Hierarchical Residuals: A Minimalist Decoupling of Residual Width from Compute Width
📄 Paper HF Daily Papers 21 ngày trước

Residual connections underpin Transformers, but they implicitly tie the residual width carried through depth to the compute width used inside each layer. We introduce Hierarchical Residuals (HiRes), a minimalist architectural change that decouples these dimensions, widening the r…

🅲 hạng C
0
ODE Policy: Robot Learning via Action Dynamics, Not Actions
📄 Paper HF Daily Papers 20 ngày trước

Most robot policies, whether trained by imitation learning (IL) or reinforcement learning (RL), map observations to action values at discrete control steps and leave the continuity of physical motion to be learned implicitly. We propose ODE policy, which learns the dynamics of th…

🅲 hạng C
0
Flide: Accelerating Decision Inference with Exact State Forking
📄 Paper HF Daily Papers 21 ngày trước

Decision models such as Jev have emerged as a highly promising paradigm for LLM agent systems, providing lightweight and reliable categorical reasoning for rapid System One execution. However, existing generation-oriented inference engines fail to support these workloads effectiv…

🅲 hạng C
0
TRIO: Task Refinement with Individual Optimization for Retrieval-Augmented Generation
📄 Paper HF Daily Papers 19 ngày trước

This paper studies retrieval-augmented generation (RAG), which leverages external knowledge to improve the performance of language models. Despite the remarkable progress, current RAG approaches are still far from satisfactory due to inadequate retrieval and potential hallucinati…

🅲 hạng C
0
AutoVend: Benchmarking LLM Agents on Long-Horizon Autonomous Vending Operations
📄 Paper HF Daily Papers 24 ngày trước

Large language model (LLM) agents are extending agent-mediated commerce on the Web from shopping for consumers to operating businesses for sellers. In online-to-offline (O2O) commerce, decisions are issued through Web-based platforms but realized through costly physical execution…

🅲 hạng C
0
Stochastic Deep Supervision for Gallery-Consistent Early-Exit Face Recognition
📄 Paper HF Daily Papers 19 ngày trước

Vision Transformers (ViTs) perform strongly in face recognition (FR), yet their fixed-depth inference limits deployment in resource-constrained settings. We propose SD-ViT, a training framework that gives a usable exit at every block depth by making every depth produce discrimina…

🅲 hạng C
0
LCM-Promo: An LLM-Based Uplift Model for Promotion Allocation
📄 Paper HF Daily Papers 24 ngày trước

We present LCM-Promo, an LLM-based uplift model for promotion allocation. LCM-Promo is a task-specific Large Commerce Model (LCM): a finetuned LLM trained on transactional commerce data to predict user intent and promotion response. It learns monotonic promotion-response curves f…

0
Auriga Technical Report: A Task-Agnostic Agent Harness for Zero-Shot Embodied Manipulation
📄 Paper HF Daily Papers 21 ngày trước

Vision-language models can interpret instructions and plan manipulation tasks, yet struggle when acting directly as robot policies. On RoboDojo, a large planning-execution gap separates high-level task decisions from physical execution: removing arm execution raises GPT-6 Astra's…

🅲 hạng C
0
Modality-Aware Expert Pruning for MoE-Based Multimodal Large Language Models
📄 Paper HF Daily Papers 17 ngày trước

Mixture-of-Experts (MoE) enables multimodal large language models (MLLMs) to scale capacity efficiently, but expert parameters dominate memory even though only a few experts are activated per token: they account for over 90% of model size in modern MoE-MLLMs (e.g., 54 GB out of a…

🅲 hạng C
0
DepthGraft: Structural Regularization through Hierarchical Cross-Layer KV Reconstruction
📄 Paper HF Daily Papers 19 ngày trước

Scaling depth is a key route to stronger large language models, but Pre-LN architectures often suffer from depth-wise information dilution, limiting the effective use of deep layers. While cross-layer connectivity has proven effective at mitigating this problem, existing designs …

🅲 hạng C