| tech.blog / monetary.live / ivan.digital
August 03, 2026 at 12:07 UTC
πŸ€— HuggingFace Daily Papers
From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM...
β–² 62 πŸŽ“ Duke University
  • Reinforcement Learning with Verifiable Rewards (RLVR) has driven recent progress in reasoning-oriented large language models (LLMs) by enabling large-...
  • However, its applicability remains largely limited to domains such as mathematics and coding, where correctness can be deterministically verified.
  • Open-ended tasks instead often rely on human preferences, reward models, or LLM-based judges, introducing evaluation bias, judge capability bottleneck...
  • RLSVR transforms open-ended tasks into verifiable proxy environments whose internal rules and interaction outcomes automatically generate reward signa...
N_0-VTLA: Scaling Vision-Tactile-Language-Action Model with Latent Tactile Tokens
β–² 45 🏒 NeoteAI
  • We present N_0-VTLA, a vision-tactile-language-action (VTLA) foundation model capable of (1) fine-grained contact-rich manipulation with tactile perce...
  • Building on current vision-based backbones, we propose a training recipe for tactile integration consisting of visuo-tactile pre-training, staged tact...
  • During pre-training, the policy learns broad contact priors from NeoData, our large-scale visuo-tactile robot dataset; to our knowledge, N_0-VTLA is t...
  • During post-training, we augment the policy with a predictive tactile pathway that distills the contact patterns learned at scale into the fine motion...
AISPA: User-Centric System Prompt Auditing for Large Language Model Applications
β–² 29 πŸŽ“ Stanford University
  • System prompts are instructions configured by developers to govern the behaviors of foundation models in AI applications.
  • They are used throughout commercial AI products, but are rarely disclosed to the public or regulators, creating a serious trust and accountability gap...
  • In this paper, we introduce Artificial Intelligence System Prompt Assurance (AISPA), a user-centric framework for systematically auditing system promp...
  • AISPA examines specific parts of a system prompt and evaluates them along eight dimensions that matter to users.
N_0-TWAM: Scaling Tactile-Native World-Action Model for Contact-Rich Manipulation
β–² 23 🏒 NeoteAI
  • We present N_0-TWAM, a tactile-native world-action model for contact-rich manipulation that predicts both future vision and future contact.
  • To our knowledge, it is the first tactile world-action model trained at large scale, and it shows strong capability on contact-rich tasks.
  • We pre-train N_0-TWAM at large scale with visuo-tactile joint training over tactile-rich demonstrations spanning six embodiments and 450 tasks.
  • We use NeoForce, a unified force-based tactile representation, to form a physically grounded contact signal that conditions action generation.
πŸ›οΈ Top Research Institutions
Tool Specifications Matter: Uncovering and Mitigating Safety Risks in AI Agents
πŸ›οΈ Beijing University of Posts and Telecommunications AI & Machine Learning
  • Large language models (LLMs) often face challenges in effectively utilizing memory for personalization.
  • The paper investigates memory utilization in LLMs and proposes methods to enhance their ability to act on relevant knowl...
RayViT: Ray-Conditioned Visual Representations for Viewpoint-Robust Imitation Learning
πŸ›οΈ Karlsruhe Institute of Technology AI & Machine Learning
  • The challenge of efficiently training multi-policy large language models (LLMs) often leads to suboptimal performance du...
  • The paper proposes an automated task sequencing method to enhance the training efficiency of multi-policy LLMs.
Rethinking AI Cloud Infrastructure for Agentic Serving Systems with the Aries Experimentat...
πŸ›οΈ NTU Singapore Systems & Infrastructure
  • Traditional AI cloud infrastructure struggles to efficiently serve autonomous agents requiring persistent context and to...
  • The Aries Experimentation Framework, designed for agentic serving systems, integrates repeated inference with sandboxed ...
The Kikuchi Hierarchy is Sharp for $k$XOR
πŸ›οΈ MIT Theory & Algorithms
  • Understanding the computational limits of the Kikuchi hierarchy in relation to the planted noisy kXOR problem is crucial...
  • The paper establishes that the Kikuchi hierarchy is sharp for the kXOR problem, demonstrating a conjectured trade-off be...
CoLAS: Multimodal Corroboration of Latent Asset Signals for Financial Trading
πŸ›οΈ National University of Singapore Other CS
  • Financial trading relies on extracting reliable signals from heterogeneous market modalities, which is challenging due t...
  • CoLAS introduces a multimodal corroboration framework that integrates diverse market signals to enhance trading decision...
Local Stochastic Rough Volatility: Pathwise Filtering and the Conditional Density Equation
πŸ›οΈ Imperial College London Quantitative Finance
  • Modeling the conditional density in local stochastic rough volatility frameworks remains complex and underexplored.
  • This paper studies the conditional-density equation and its pathwise transformation in local stochastic rough volatility...
Linear Estimation of Structural and Causal Effects for Nonseparable Panel Data
πŸ›οΈ MIT Economics
  • Estimating structural and causal parameters in nonseparable models using panel data is complex due to unobserved heterog...
  • Development of linear estimators that account for time-varying individual heterogeneity in panel data.
AI & Machine Learning
321 papers 7 cats
Systems & Infrastructure
27 papers 6 cats
Software & Programming
29 papers 4 cats
Theory & Algorithms
25 papers 5 cats
Applications
92 papers 7 cats
Other CS
24 papers 9 cats
Quantitative Finance
20 papers 9 cats
Economics
17 papers 3 cats