… The following papers were recommended by the Semantic Scholar API OSExpert: Computer-Use Agents Learning Professional Skills via Exploration 2026 EnterpriseOps-Gym: Environments and Evaluations for Stateful Agentic Planning and Tool Use in Enterprise Settings 2026 OccuBench: Evaluating AI Agents on… …
… We partition each trajectory with 30-min windows and release two datasets: Real ICU -Gold with 930-window annotations from 94 MIMIC-IV patients, and Real ICU -Scale with 11,862 windows extended by Oracle, a physician-validated LLM hindsight labeler. …
… AI-generated summary Workspace learning requires AI agents to identify, reason over, exploit, and update explicit and implicit dependencies among heterogeneous files in a worker's workspace, enabling them to complete both routine and advanced tasks effectively. …
… These challenges are particularly pronounced for open-source LLMs with smaller parameter sizes, limited context windows, and constrained inference budgets, which contribute to increased error accumulation in agentic settings. …
… This leads to two challenges: partial observability requires maintaining uncertainty over unobserved world attributes, and long interaction history causes context to grow without bound, diluting task-relevant information. …
… AI-generated summary Does a lexical retriever suffice as large language models LLMs become more capable in an agentic loop ? This question naturally arises when building deep research systems. We revisit it by pairing BM25 with frontier LLMs that have better reasoning and tool-use abilities. …
… AFTraj-2K dataset — We construct AFTraj-2K, a curated corpus of agentic trajectories spanning Coding, Math, and Agentic domains, pairing strictly filtered safe runs with multi-judge verified failure runs annotated at their decisive error step A compact online auditor — We develop AgentForesight -7B… …
… The following papers were recommended by the Semantic Scholar API OpenSeeker: Democratizing Frontier Search Agents by Fully Open-Sourcing Training Data 2026 Marco DeepResearch: Unlocking Efficient Deep Research Agents via Verification-Centric Design 2026 DR-Venus: Towards Frontier Edge-Scale Deep R… …
… AI-generated summary On-policy distillation OPD is widely used for LLM post-training. When pushed with a reward-extrapolation coefficient lambda 1, the student can lift past the teacher in domain, but past a threshold lambda the same step violates the output contract on structured-output tasks . …
… View arXiv page View PDF Project page Add to collection Community This comment has been hidden the latent fusion across overlapping diffusion windows to preserve global consistency while conditioning on arbitrary segment maps is the part that sticks with me. i'd like to see an ablation where you di… …