AI research papers.

Video generation begins in text space by authoring a cinematic screenplay, then materializes into pixels.

Existing point tracking models face a fundamental tradeoff: they can either track a sparse set of query points over long horizons, or track all points across only short clips.

Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discrete decisions are tightly coupled to geometric, kinematic, and dynamic constraints.

World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation.

Retrieval-based factuality evaluation, where LLM-generated claims are verified against evidence from authoritative medical corpora, has become the dominant paradigm for scalable hallucination detection in high-stakes cli…

Optimizing embedding systems for production recommendation pipelines demands systematic exploration that consumes disproportionate engineering effort at scale.

Robots are getting smarter, but how can their hardware match that growth?

MentalHealthBench is an expert-informed benchmark for evaluating helpful and safe AI responses across realistic mental health conversations.

Scalable simulation is essential for robot data generation, policy training, evaluation, and safe iteration, yet real-world interaction is costly and conventional simulators require labor-intensive construction.

Self-attention-based vision backbones perform well on dense prediction, but the quadratic computational cost of global softmax attention limits their efficiency as image resolution increases.

Time series agents answer analytical questions by calling external tools, and which tools they carry is decided by people before the agent runs.

We present Hunyuan-A13B, an open-source large language model based on a Mixture-of-Experts architecture.

We study coding agents for long-horizon, dexterous robotics and ask whether their solutions can provide scalable supervision for learning general robot policies.

Language-model agents increasingly face long-horizon tasks with evolving state, interdependent decisions, and delayed outcomes.

Agentic memory systems reuse past experience to improve future performance, yet most existing designs curate memory at write time: once a task is completed, its trajectory is distilled into a fixed artifact, such as a re…

AI research agents need reliable knowledge of how their experiments change outcomes.

Physical intelligence requires more than predicting how the world may evolve: predictions must remain actionable as the world continues to change.

Solutions based on large language models (LLMs) often rely on temperature sampling to improve accuracy and stability by aggregating multiple samples from the completion distribution.

Video is a rich representation of a physical event, capturing appearance, geometry, motion, and temporal evolution.

Pruning large pre-trained transformer-based ASR models such as OpenAI's Whisper has seen great adoption, as pruning the decoder led to significant end-to-end transcription speedups.

Long-horizon embodied interaction requires agents to retain and continually update information about the environment as they observe, act, and encounter change.

Vision-Language-Action models provide a strong foundation for general-purpose robot control, yet a vast majority of policies do not preserve and leverage episode-level information beyond the current observation.

Pretrained generative Diffusion Transformers (DiTs) capture rich pixel-level visual and language-conditioned structure through large-scale image and video generation training.

Recent advances in large language models (LLMs) have enabled agents to tackle long-horizon tasks across diverse environments.

Advances in generative models have improved video fidelity, enabling long-horizon generation, interactive world modeling, and evolving visual environments.

Artificial intelligence offers an unprecedented opportunity to augment human capabilities, yet progress at the frontier has focused primarily on advancing model capabilities.

Representation Autoencoders (RAEs) enable diffusion models to operate in the feature spaces of pretrained visual encoders.

Recently, Large Language Models (LLMs) have been increasingly able to solve advanced mathematical problems, including many that have been open for decades.

Object permanence and solidity are hallmarks of human cognitive priors.

Human hand-object interactions (HOIs) provide a rich source of demonstrations for dexterous manipulation, but learning directly from them presents challenges in bridging morphology gaps, ensuring dynamical feasibility, a…

World-action models (WAMs) transfer visual and motion priors from pretrained video generators to robot control by jointly modeling visual dynamics and actions.

On-policy distillation (OPD) corrects a student on the responses it writes, but its signal is the teacher's next-token distribution: it tells the student what the teacher says but misses how it thinks.

arXiv:2606.18193v2 Announce Type: replace-cross Abstract: We evaluate the adversarial robustness of three frontier large language models (LLMs) developed by Anthropic, Opus 4.8, Fable 5 and its successor Fable 5.1, again…

arXiv:2609.24246v1 Announce Type: cross Abstract: Large language models (LLMs) have shown remarkable progress in natural language understanding, yet their effectiveness in specialized fields like astronomy and astrodynam…

arXiv:2609.24698v1 Announce Type: cross Abstract: Repeated execution of the target model during autoregressive decoding is a major source of LLM inference latency.

Modern embodied agents achieve impressive success rates, yet their actual instruction-following ability is far weaker than these numbers suggest.

Reference-based image quality assessment (IQA) metrics aim to reflect how humans perceive the perceptual distance between a pair of images.

LLM agents increasingly work on long-horizon tasks, and the decisions they make along the way, such as which hypothesis to test or which implementation to build on, determine the outcome of the whole run.

Learned-memory methods store information in an explicit table and consume it through a separate reader, allowing addressing, storage, and reading to be modified independently.

Transformer architectures exhibit cross-layer redundancies, yet post-training compression pipelines typically optimize layers in isolation or rely on heuristic grouping strategies that disregard layer-specific activation…

Prefill and decode reward different approaches to quantization: low-precision arithmetic accelerates prompt processing, while compact weights reduce memory traffic during generation.

Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematical definition, leaving its relationship to commonly used trai…

AI agents are beginning to automate research and development across the AI stack, from improving training efficiency to optimizing inference.

Calibration of language models -- the alignment between expressed or implicit confidence and empirical correctness -- is a well-studied subfield within NLP.

LLM-as-a-judge enables evaluation across diverse tasks, but inference cost and confidence reliability become critical at scale.

We introduce Knowledge Pull Requests (KPRs), a framework for continual document authoring that makes each change interpretable.

The rapid capability gains of frontier language models are widely attributed to improved reasoning abilities, yet this cannot be verified as raw CoT traces in closed-source systems are hidden.

Vector Quantization (VQ) is fundamental to discrete visual tokenizers that power modern autoregressive and masked image generation models.

Long-term conversational memory in multi-party settings requires more than retrieving relevant content from long-term conversations: it must distinguish who said what, whom each statement concerns, how individuals percei…

A multi-agent system can reduce latency on complex tasks by executing work concurrently.