PVPO: Pre-Estimated Value-Based Policy Optimization for Agentic Reasoning

Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback

UQ: Assessing Language Models on Unsolved Questions

CARJAN: Agent-Based Generation and Simulation of Traffic Scenarios with AJAN

TiKMiX: Take Data Influence into Dynamic Mixture for Language Model Pre-training

TalkVid: A Large-Scale Diversified Dataset for Audio-Driven Talking Head Synthesis

Droplet3D: Commonsense Priors from Videos Facilitate 3D Generation

A.S.E: A Repository-Level Benchmark for Evaluating Security in AI-Generated Code

EmbodiedOneVision: Interleaved Vision-Text-Action Pretraining for General Robot Control

R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning

Igniting Creative Writing in Small Language Models: LLM-as-a-Judge versus Multi-Agent Refined Rewards

TMUAD: Enhancing Logical Capabilities in Unified Anomaly Detection Models with a Text Memory Bank

Analysing Chain of Thought Dynamics: Active Guidance or Unfaithful Post-hoc Rationalisation?

AWorld: Orchestrating the Training Recipe for Agentic AI

MCP-Bench: Benchmarking Tool-Using LLM Agents with Complex Real-World Tasks via MCP Servers

rStar2-Agent: Agentic Reasoning Technical Report

Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning

MobileCLIP2: Improving Multi-Modal Reinforced Training

AI-AI Esthetic Collaboration with Explicit Semiotic Awareness and Emergent Grammar Development

Gaze into the Heart: A Multi-View Video Dataset for rPPG and Health Biomarkers Estimation

Predicting the Order of Upcoming Tokens Improves Language Modeling

MIDAS: Multimodal Interactive Digital-human Synthesis via Real-time Autoregressive Video Generation

Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policies

Self-Rewarding Vision-Language Model via Reasoning Decomposition

Beyond Transcription: Mechanistic Interpretability in ASR

CODA: Coordinating the Cerebrum and Cerebellum for a Dual-Brain Computer Use Agent with Decoupled Reinforcement Learning

WebSight: A Vision-First Architecture for Robust Web Agents

UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning

Hermes 4 Technical Report

OmniHuman-1.5: Instilling an Active Mind in Avatars via Cognitive Simulation

PVPO: Pre-Estimated Value-Based Policy Optimization for Agentic Reasoning

Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback

UQ: Assessing Language Models on Unsolved Questions

CARJAN: Agent-Based Generation and Simulation of Traffic Scenarios with AJAN

TiKMiX: Take Data Influence into Dynamic Mixture for Language Model Pre-training

TalkVid: A Large-Scale Diversified Dataset for Audio-Driven Talking Head Synthesis

Droplet3D: Commonsense Priors from Videos Facilitate 3D Generation

A.S.E: A Repository-Level Benchmark for Evaluating Security in AI-Generated Code

EmbodiedOneVision: Interleaved Vision-Text-Action Pretraining for General Robot Control

R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning

Igniting Creative Writing in Small Language Models: LLM-as-a-Judge versus Multi-Agent Refined Rewards

TMUAD: Enhancing Logical Capabilities in Unified Anomaly Detection Models with a Text Memory Bank

Analysing Chain of Thought Dynamics: Active Guidance or Unfaithful Post-hoc Rationalisation?

AWorld: Orchestrating the Training Recipe for Agentic AI

MCP-Bench: Benchmarking Tool-Using LLM Agents with Complex Real-World Tasks via MCP Servers

rStar2-Agent: Agentic Reasoning Technical Report

Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning

MobileCLIP2: Improving Multi-Modal Reinforced Training

AI-AI Esthetic Collaboration with Explicit Semiotic Awareness and Emergent Grammar Development

Gaze into the Heart: A Multi-View Video Dataset for rPPG and Health Biomarkers Estimation

Predicting the Order of Upcoming Tokens Improves Language Modeling

MIDAS: Multimodal Interactive Digital-human Synthesis via Real-time Autoregressive Video Generation

Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policies

Self-Rewarding Vision-Language Model via Reasoning Decomposition

Beyond Transcription: Mechanistic Interpretability in ASR

CODA: Coordinating the Cerebrum and Cerebellum for a Dual-Brain Computer Use Agent with Decoupled Reinforcement Learning

WebSight: A Vision-First Architecture for Robust Web Agents

UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning

Hermes 4 Technical Report

OmniHuman-1.5: Instilling an Active Mind in Avatars via Cognitive Simulation