Latest AI/ML research from ArXiv and Papers with Code
Financial institutions operating under data-residency constraints are restricted to open-weight models small enough to locally host, which trail frontier systems on multi-step quan...
AI agents must decide when another search is worth its cost. Answer-success probabilities capture question difficulty, while this decision requires estimating the additional benefi...
Multi-agent systems (MAS) require both effective coordination and specialized task execution. Existing prompt optimization approaches typically couple these functions within agent-...
Safe household navigation requires robots to regulate clearance and speed according to what an obstacle is and how close it is, without unnecessary conservatism. Yet existing evalu...
Multimodal intelligence can be viewed as human-guided co-evolution between worlds and agents through interaction. However, scaling this process faces two bottlenecks: constructing ...
Long-horizon agents continuously encounter reusable procedural knowledge, yet existing systems typically rely either on model adaptation or external memory, trading persistent lear...
Are concepts and mechanisms unified across modalities in current AI models, or instead implemented separately for each one? We operationalize this broad question of modality unific...
Diffusion language models (DLMs) promise parallel text generation, but in practice face an unresolved speed-quality trade-off. Exact decoding fills in only one masked position per ...
Action-conditioned world models offer a scalable alternative to costly real-world interaction for improving robot policies. However, existing world-model-based methods largely trea...
Agent harnesses enable large language models to act through tools, observe the results, and iteratively work toward a task goal. Verification checks and revises agents' work to imp...
Animals have a remarkable capacity for autonomy---driven by intrinsic motivation---without external sensory cues or extrinsic objectives, yet the computational mechanisms behind th...
Scaling vision-language-action (VLA) models is constrained by scarce action-labeled demonstrations and incompatible control spaces across robot embodiments. We present RynnVLA-Late...
Multi-agent systems often rely on predefined collaboration structures, limiting their ability to adapt to diverse task requirements. We introduce DynaFlow, a framework that integra...
In this paper, we introduce Δτ-Bench, an executable agent benchmark that models the structural evolution of large tool ecosystems. In Δτ-Bench, independently evolving services cont...
Reusable Web-agent verification requires a rule to follow changing requirements and support subsequent execution. We introduce AgentProofBench, which connects paired tests of a fro...
Reinforcement learning (RL) has emerged as the primary paradigm for optimizing Large Language Model (LLM) agents, yet efficient critic-free algorithms (e.g., GRPO) suffer from a fu...
While progress in informal mathematics has come largely from scaling the chain-of-thought and agentic reasoning of frontier language models, formal theorem proving has increasingly...
Memory is increasingly used in long-horizon agents, yet the net value of building memory over direct retrieval from raw history remains largely unclear. Through a large-scale evalu...
Extending Vision-Language-Action models (VLAs) to humanoid robots requires coordinated whole-body control and diverse demonstrations that are costly to collect in real-world scenar...
Jailbreak benchmarks test one checkpoint and treat which jailbreaks break it as a property of its pretraining data and safety training. We ask whether another pretraining run of th...
People have long dreamed of an assistant that is always there to guide them and remind what they forget, from new dishes to lost keys. Powered by multimodal large language models (...
Continual learning is the ability of LLM agents to improve over time by acquiring, refining, and applying knowledge from experience while retaining useful prior knowledge. It is es...
Recursive self-improvement (RSI) requires an agentic system to design, train, and evaluate its own successor. Data-driven post-training is a central part of that loop, yet existing...
Spatial audio authoring coordinates source decisions with soundfield synthesis. For editing, plans specify what to change and retain, while reference audio supplies further acousti...