Tool use, function calling, orchestration and the protocols that let models act rather than only answer.
arXiv:2610.27474v1 Announce Type: new Abstract: The represented gradient flow of a parameterized model depends on a mobility that can retain hidden parameter information. We study when one entire rep...
arXiv:2610.27623v1 Announce Type: new Abstract: Reinforcement learning for long-horizon language agents requires identifying whether intermediate actions make eventual task success more or less likel...
arXiv:2610.28565v1 Announce Type: new Abstract: Clustered federated learning (CFL) can suffer from cluster collapse. In particular, under unfavorable initialization, early client assignments concentr...
arXiv:2610.27100v1 Announce Type: cross Abstract: Realistic market impact is essential for generative limit order book (LOB) models to serve as world models for trading agents. If a simulator mispric...
arXiv:2610.27141v1 Announce Type: cross Abstract: Reliable contact establishment remains a critical challenge for autonomous on-orbit servicing (OOS), particularly when interacting with non-cooperati...
arXiv:2610.28315v1 Announce Type: cross Abstract: Large language model (LLM) agents increasingly use persistent memory to personalize responses, and personal attributes in this memory pose a privacy...
arXiv:2610.28004v1 Announce Type: new Abstract: Camera design heavily influences perception task performance, yet cameras are typically designed by domain experts through costly experimentation, in i...
arXiv:2610.28346v1 Announce Type: new Abstract: Autonomous video editing from open-world resources remains a challenging problem in multimedia content creation. Retrieval-based methods are constraine...
arXiv:2610.28630v1 Announce Type: new Abstract: Game engines and coding agents can specify the layout, motion, and camera of a scene, yet turning such coarse structure into realistic video requires a...
arXiv:2601.07823v2 Announce Type: cross Abstract: Video world models have emerged as promising candidates for high-fidelity world models, offering the potential to synthesize high-quality videos capt...
arXiv:2605.20682v3 Announce Type: replace Abstract: Small defects in unfamiliar product categories can be difficult to distinguish from benign appearance variations in a global view. Additional reaso...
MAS-Guard is a unified framework designed to safeguard both single- and multi-agent large language model (LLM) systems. It introduces a local Agent Guardrail (A-Guard) layer that uses task-specific experts to evaluate agentic events, and a system-level Control Plane that applies deterministic structural checks and semantic assessments to routes, capabilities, and message content. The framework enforces interventions—allow, reflection, confirmation, and block—across benchmarks, reducing attack success rates by 71.5‑86.3% while maintaining utility on benign tasks.
Memory as Code (MaC) introduces a unified memory framework that integrates memory construction and retrieval via a shared code abstraction. It builds a three‑layer executable memory structure—atomic functions, relational links, and modular compositions—and uses declarative retrieval to translate queries into structured programs. Experiments on LoCoMo and LongMemEval demonstrate up to 15.91% higher LLM‑judge accuracy and reduced token usage compared to leading baselines.
SciDiagBench is a new benchmark designed to evaluate scientific diagram generation in a way that mirrors human peer‑review practices. It provides a dataset of 165 diagrams from 10 domains, systematically classified into 11 distinct types such as graphical abstracts and technical roadmaps. The benchmark also introduces a preference‑guided calibration method that incorporates human preferences into type‑specific evaluation criteria, aiming to improve the reliability of automated assessments.
Navalia is a framework that advances cowork agents by combining verifiable task synthesis with long‑horizon post‑training. It selects supporting evidence from corporate financial reports, builds executable computation chains, and uses independent LLM review and programmatic checks to ensure evidence consistency and answer correctness. The verified tasks are then used for trajectory collection, supervised fine‑tuning, and reinforcement learning, yielding consistent performance gains over the base model across five financial and office‑task benchmarks.
ReuF2F is a Lean 4 benchmark comprising 244 formalized open problems sourced from Research Experiences for Undergraduates (REU). It aims to evaluate coding agents on less prominent, entry‑level research problems across diverse mathematical subfields, using a Docker‑based Lean project setup and a unified harness called Leanimum‑agent. The study finds that even the best model solves only about 40 % of the problems, indicating significant room for improvement.
CovNav is a training‑free framework that uses Conformal Prediction to decide when an embodied agent should switch from fast, low‑latency LLM reasoning to slower, more accurate chain‑of‑thought reasoning during object‑goal navigation. By treating each spatial query as a binary decision and calibrating the fast LLM’s token probabilities, CovNav escalates only ambiguous queries to the slow mode, achieving higher success rates on Gibson and HM3D while invoking slow reasoning for just 12.4% of queries. Real‑robot tests on a TurtleBot 4 confirm the method’s practical benefits.
The paper introduces a geo‑referenced smartphone dashcam dataset of 63,839 frames from South Texas and a baseline framework for assessing road lane quality. Three new metrics—Contrast Stability Index (CSI), Line Continuity Ratio (LCR), and Regression Lane Confidence Score (RLCS)—are proposed to evaluate lane‑marking visibility, continuity, and map consistency without dense human labels. Baseline models achieve strong performance, and pilot validation shows the metrics correlate better with human judgments than generic image‑quality measures.
Object-Oriented Harness (OO-Harness) is a framework that optimizes agent harnesses by evolving a hierarchy of executable harness classes through object-oriented design. During training, an evolution agent uses execution records to create specialized harness subclasses, while at test time a harness router selects the appropriate harness for each task based on learned routing lessons. Experiments on three long-horizon benchmarks show OO-Harness outperforms other methods, achieving up to a 15% absolute improvement over the base harness.
The paper introduces cache engineering for large language model agents, enabling direct control over the active key-value cache without altering the trajectory or triggering costly re-prefilling. It defines cache operations such as drop and repos, formalizes their effects, and addresses the challenge of history-dependent cache states. The authors co-design a stateless AKV serving engine that manages stateful cache reuse, eviction, and restoration, achieving significant performance and cost improvements over traditional summarization methods.