AI research papers.

In a small language model the vocabulary is the single largest part.

Spatio-Temporal Video Grounding (STVG) aims to jointly localize a target in space and time given a natural-language query.

Enhancing the agentic capabilities of Large Language Models (LLMs) to tackle complex software engineering (SWE) and general terminal-usage problems is a pressing challenge.

Harness evolution automatically optimizes the prompts, skills, tools, and orchestration code around a fixed-weight model, offering a way to improve agentic systems without updating model weights.

GPT-6-Astra demonstrates a level of competence in robotic manipulation and navigation that challenges the conventional paradigm of building specialized models for individual tasks and embodiments.

Command-line interface (CLI) agents increasingly execute longer-horizon tasks, yet existing benchmarks either assume that user instructions remain fixed throughout execution or evaluate task outcomes without capturing ru…

On-policy self-distillation (OPSD) enables large language models (LLMs) to learn from privileged information (PI) without a separate teacher model.

Retrieval is a fundamental task in multimodal systems: given an image or text query, the goal is to find the most relevant images or texts in a collection, often using a shared embedding space.

Large language models (LLMs) are increasingly used in critical domains, where their susceptibility to hallucinations makes reliable and interpretable uncertainty quantification essential.

Methods for training language models on sensitive data differ in how they balance capability gains against the risk of leaking private information.

Forecasting an unresolved event requires finding relevant evidence and deciding how strongly it supports each outcome.

AI agents increasingly browse the web on users' behalf, yet they consume webpages as screenshots or text, exposing their LLMs to unvetted third-party content.

Text-based world models have recently shown potential to simulate complex RL environment state transitions for a diverse set of use cases.

Augmenting softmax attention with forget gates typically relies on summing decay contributions along the key--query path.

Split language models keep input text on a client but send intermediate representations to a server.

Advances in frontier coding agents have sharply increased the token costs of AI-assisted software engineering.

Creative intent in video editing can be expressed through text instructions (e.g., "remove the people"), image references (e.g., subject appearance), video references (e.g., camera movement), or a combination of the abov…

We introduce Continuum Attention (CA), a causal Transformer attention operator that replaces discrete key and value lookup with integration over a continuous, piecewise linear memory curve.

Multi-agent systems allow language models to solve tasks collaboratively by sharing complementary information.

In long-horizon tool-calling tasks, LLM agents can fail not only because of tool-use errors, but also because of ineffective interaction with users.

Verifiable instruction-following benchmarks emphasize output-constraint satisfaction but can overlook whether responses meaningfully answer the underlying request, creating an evaluation blind spot.

Multimodal models and agents have emerged to enable AI systems to learn from and interact with the real world more like humans by integrating information across multiple modalities.

Vision-language-action (VLA) models have rapidly advanced generalist robot manipulation by harnessing the semantic knowledge of pretrained vision-language backbones.

Iterative sampling in diffusion models can lead to error accumulation due to exposure bias.

Slider-based control aims to give users continuous, fine-grained control over the intensity of individual edits in instruction-based image editing.

Can vision-language models turn 2D observations into 3D understanding that supports precise action?

Music source separation (MSS) has been conventionally addressed by discriminative methods.

Embodied visual tracking requires a robot to identify a designated target from visual observations and control its own motion to follow that target through an environment.

Language-model agents that run long tasks often redo work they have already done: they re-issue a tool call they made earlier and get back exactly the same observation, a step that adds no new information.

We seek a hierarchical resolution of residual error under a composite learning objective as a principled mechanism for identifying and exploiting additional information, particularly under distribution shift and perturba…

Long-horizon agentic RL is constrained by the futures visited by the current policy.

MLLM-based detectors formulate vision-language grounding as autoregressive coordinate generation, yet in dense scenes, they often collapse into a box repetition loop, repeatedly emitting identical boxes until the decodin…

Simulated environments enable scalable training data generation for embodied agents, but more data does not necessarily provide what a particular agent needs to learn.

Large language model (LLM) agents accomplish complex tasks through multi-turn interactions with external tools.

Agentic capabilities are increasingly important for modern LLMs, yet remain difficult to evaluate during pretraining.

Comparing financial disclosures across languages requires identifying shared accounting concepts and recovering all numerical facts that express them.

Text-to-image models increasingly generate readable Chinese text that still contains local rendering defects, such as a character with an extra or missing stroke or a malformed component.

Repeated event forecasting requires extracting predictive signals from noisy evidence and evaluating their value across successive decisions.

We propose DAGFormer, a Transformer language model whose cross-layer connections are planned by a jointly trained topology predictor.

Recent work has explored visual reasoning through image generation, where models solve visual-reasoning problems by directly generating solution images.

Large language models (LLMs) hold broad knowledge in one set of weights, but lack interpretability and trustworthy reasoning, especially in high-stakes use cases.

Large language model agents increasingly undertake long-horizon tasks in complex environments characterized by fragmented evidence, partially completed states, and dependencies across resources and artifacts.

Reinforcement Learning with Verifiable Rewards (RLVR) improves reasoning in language models, but scalar outcomes provide coarse supervision over response trajectories.

Large language models (LLMs) increasingly tackle complex real-world tasks in specialized domains.

We present a Generative World Simulation framework that couples persistent shared world state with continuous first-person audiovisual generation.

Monocular world-space human recovery methods excel on current benchmarks, but these rarely include fast camera motion.

Self-distillation in agentic reinforcement learning has produced a growing number of methods.

Distilling capable language-model agents into smaller models requires aligning the student's behavior with the teacher's.

We propose DeltaStreamer, a novel agentic framework for streaming video understanding inspired by how humans observe, remember, and investigate.

Looped Mixture-of-Experts (MoE) models combine parameter reuse with sparse computation to improve performance under matched budgets.