arXiv:2607. 05804v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student policy by matching a stronger teacher on the student's own trajectories, offering a promising framework for language agent training.
arXiv:2607. 23765v1 Announce Type: cross Abstract: Large language models (LLMs) achieve impressive performance across multiple domains, but using the most capable model for every query is prohibitive at scale.
Building capable visual web agents requires long-horizon reasoning, precise grounding, and robust interaction with dynamic real-world websites. Despite rapid progress, the strongest systems remain largely proprietary, while open agents still depend heavily on supervised post-training over large collections of curated web trajectories.
SCAFFOLD is a self‑improving framework for visual web agents that automatically induces parametric, executable skills from successful trajectories and organizes them into a recursively composed hierarchy. It compresses the skill library using a minimum‑description‑length criterion and behavioral equivalence checks, and periodically distills these skills back into model weights to internalize the abstractions. Experiments on WebArena, VisualWebArena, and Online‑Mind2Web show that SCAFFOLD raises success rates by 11.1–17.2 absolute points over the best skill‑augmented baseline and continues to improve across five self‑improvement iterations without collapsing the library.
arXiv:2606. 02031v1 Announce Type: cross Abstract: Building capable visual web agents requires long-horizon reasoning, precise grounding, and robust interaction with dynamic real-world websites.
arXiv:2609.14138v1 Announce Type: cross Abstract: As LLM agents become integrated into increasingly complex workflows, they must continually acquire new capabilities while retaining competence on pre...
arXiv:2506. 01952v2 Announce Type: replace-cross Abstract: Powered by large language models (LLMs), web browsing agents operate graphical user interfaces in a human-like manner, offering a transparent and general framework for automating web-based tasks.
We study on-policy distillation (OPD) for agentic tasks, where an LLM agent interacts with an environment over multiple turns and a student imitates a teacher over these multi-turn interaction histories. Fully online OPD is costly because each update requires fresh student rollouts through the environment and teacher queries at visited histories.
arXiv:2511. 12997v2 Announce Type: replace Abstract: Multimodal LLM-powered agents have recently demonstrated impressive capabilities in web navigation, enabling agents to complete complex browsing tasks across diverse domains.
Naive Prompt Optimization (NPO) is a lightweight, single‑lineage method that iteratively refines prompts using a teacher model’s rollout feedback. It matches or surpasses the performance of more complex optimizers like GEPA while requiring fewer rollouts, and its advantage grows with stronger teacher models. In interactive games, NPO remains competitive, and prompts optimized by NPO transfer well to other student models within the same family.
arXiv:2606. 27814v4 Announce Type: replace Abstract: Training small language-model agents for long-horizon interactive tasks requires both fast imitation and reward-driven improvement.
CORAL is an LLM‑native harness that automates continual optimization of production recommender systems. It operates in a closed loop: an agent observes system signals, reasons over past decisions, and uses tools—including a numerical optimizer—to reconfigure the recommender while staying within a fixed operating budget. In A/B experiments on two large social platforms, CORAL improved engagement without extra serving cost on one platform and reduced serving cost without harming engagement on the other, demonstrating that a single agentic loop can replace manual engineering for ongoing system tuning.