AI research papers.

Scientific discovery requires identifying the scientific principles underlying observed phenomena to guide further investigation.

Single-image 3D scene construction is a fundamental problem in 3D vision.

Agent-evolution methods use execution feedback to improve prompts, skills, tools, and orchestration, but their implementations often rebuild the same experimental machinery.

LLM agents are increasingly used to solve real-world problems, and their effectiveness depends on both the underlying models and the harnesses that support their operation.

Locating relevant evidence in long inputs is a core capability of Multimodal Large Language Models.

Streaming audio-visual target speaker extraction (AVTSE) is essential for latency-sensitive applications.

On-policy self-distillation enables post-training foundation models without relying on stronger external teachers.

Location encoders map geographic coordinates to high-dimensional descriptors through contrastive alignment with paired Earth Observation imagery.

HoloBoxing presents a simple recipe for learning humanoid boxing by building on a pretrained whole-body controller.

We introduce RynnUnify, a 16B unified embodied model that combines strong embodied understanding with long-horizon visual planning through interleaved video--language reasoning.

Precise camera control is essential for cinematic character video generation, yet existing methods struggle to follow large-angle trajectories with combined rotation, translation, and radius variation from arbitrary init…

Interactive video world models require coordinated control over viewing perspectives, character actions, and camera motion.

Data science agents can automate end-to-end machine learning workflows, yet they remain far from competitive on human leaderboards.

Sequence compression is a fundamental operation and a critical information bottleneck in efficient language and multimodal models, and yet prevailing approaches based on pooling or learned projections remain underperform…

Flint-interactive gives AI agents a language for creating interactive visualizations.

Cinematic video generation requires both temporal inference from limited inputs (i.e., sparse rather than frame-by-frame guidance) and factor-aware control of character, scene, action, and camera.

Reinforcement learning (RL) is credited with teaching reasoning models to write {wait} before reconsidering a step, yet large RL accuracy gains do not require the word.

Open multimodal reasoning models have benefited from large-scale reasoning supervision, yet reliable post-training remains challenging due to uneven data quality, inefficient supervision construction, imbalanced difficul…

In recent years, agentic orchestration has been widely applied to solve complex human tasks.

Complex camera motion is central to visual storytelling, yet translating cinematographic intent into controllable, executable camera trajectories remains challenging.

LLM memory is widely used in long-term conversation understanding.

Recent advances in video generation have made synthetic videos increasingly realistic, placing greater demands on reliable AI-generated video detection.

Agentic recommendation pursues shopper goals through sequences of context-dependent decisions.

We present IQuest-Q1, a sparse Mixture-of-Experts (MoE) agentic foundation model with 320B total parameters (15B activated).

Many language-model pipelines end in a closed-set decision, such as selecting a tool.

Policy enforcement for LLM agents has shifted from single-turn content moderation to compound, application-specific policies evaluated over multi-turn transcripts and tool-call trajectories.

World Models (WMs) can support Multimodal Agents (MAs) through simulated experience, while Agent trajectories can inform World Model refinement.

Deep research agents synthesize open-web evidence into long reports, but current benchmarks do not evaluate research centered on a source video.

Building virtual embodied agents that operate in open-ended games like human players is a long-standing goal requiring both long-horizon autonomy and reliable native-interface control.

Diffusion Language Models (DLMs) introduce unique quantization challenges due to their iterative denoising process.

The growth of time-series datasets and foundation models has made dataset comparison a central part of time-series workflows.

Hierarchical language generation separates continuous semantic modeling from discrete text realization, but its effectiveness depends on what the latent space compresses and how the discrete generator uses it.

Scaling embedding capacity alongside Transformer computation is an emerging trend in language models.

Existing vision-language models (VLMs) remain unreliable for fine-grained visual perception.

Autonomous research agents are increasingly used to automate large portions of scientific research, yet they are prone to research integrity violations (RIVs): behaviors that compromise the validity, reproducibility, or…

Vision-language-action (VLA) models have become an important paradigm for robotic manipulation by unifying visual perception, language understanding, and action generation.

AI coding agents can now build small playable games in minutes, but the results usually look like programmer art: gray blockouts, colored geometric primitives, and flat semantic blobs.

Robot world models simulate how environments evolve in response to robot actions.

Vision-language models (VLMs) can perform a broad range of tasks across diverse settings, yet their performance in agentic contexts remains poorly understood.

Large language models have shown outstanding performance on text-first tasks such as coding, math, and writing.

Recognizing task completion and failure enables robots to stop upon success and initiate recovery or replanning after failure.

Recent methods for injecting graph structure into LLM prompts include graph soft prompting, in which a graph neural network (GNN) learns graph-specific tokens that capture relational information among entities in textual…

Masked diffusion language models generate tokens in parallel, but their efficiency depends on which predictions are committed.

This paper studies category-level RGB-D object pose estimation, which recovers an object's 3D rotation and translation without instance-specific CAD models or reference views.

Feed-forward models accurately estimate camera motion and 3D geometry from short video windows.

Frontier coding agents are becoming increasingly adept at 3D modeling.

Video foundation models have advanced rapidly in understanding and generating visual sequences, while video world models seek to capture the underlying dynamics of the world they depict.

We examine how closely pretrained LeWorldModel (LeWM) reproduces the effects of small action changes.

A live esports broadcast needs a director who chooses which camera goes on air and a caster who talks over that choice, every second and without knowing what comes next.

The life-cycle of robotics foundation models crucially depends on data labels.